| *** dailylinux has quit IRC | 00:00 | |
| *** lzyeval has joined #openstack | 00:01 | |
| *** cloudfly has joined #openstack | 00:05 | |
| *** andreas__ has quit IRC | 00:06 | |
| andrewbogott | maplebed: I've never seen it work, but in theory configuring keystone is easy, just use keystone-manage add, etc. And then configure swift with swift_auth_server=keystone | 00:07 |
|---|---|---|
| andrewbogott | (I'm guessing here :) ) | 00:07 |
| maplebed | I'm not sure I actually want keystone... | 00:07 |
| *** whit has quit IRC | 00:07 | |
| andrewbogott | Me neither | 00:07 |
| cloudfly | well it's got pluggable modules for auth mechanisms | 00:08 |
| cloudfly | that's pretty nice | 00:08 |
| *** qazwsx|2 has quit IRC | 00:08 | |
| maplebed | yeah - my current setup is using tmpauth. | 00:10 |
| *** rnorwood has joined #openstack | 00:10 | |
| notmyname | maplebed: I can try to clarify the issue | 00:14 |
| maplebed | notmyname: that would be fantastic! | 00:14 |
| notmyname | maplebed: what is still confusing? | 00:14 |
| maplebed | I followed the admin docs while setting up my swift cluster. In those docs, it instructs that I test that it's working by creating an account/user/pass and passing that in with the X-Storage-User header. | 00:16 |
| notmyname | hmm..that doesn't sound right | 00:16 |
| notmyname | ah. nm. my mistake | 00:16 |
| *** lloydde has joined #openstack | 00:16 | |
| notmyname | your are using tempauth? | 00:16 |
| maplebed | A colleague is trying to use the API docs, which all refer to a user and api key, passed in with the X-Auth-User and X-Auth-Key headers. | 00:16 |
| maplebed | I am using tempath (at the moment) | 00:16 |
| maplebed | I am confused about the relationship between the x-storage and x-auth stuff, and haven't found how to create an api key (so can't use the api docs). | 00:17 |
| maplebed | for reference, the docs I was following and am reading are: | 00:17 |
| notmyname | the username (X-Auth-User) is the tempauth account:user combo. the api key (X-Auth-Key) is the tempauth password | 00:18 |
| maplebed | http://swift.openstack.org/howto_installmultinode.html#troubleshooting-notes, which seems to overlap pretty well with http://docs.openstack.org/diablo/openstack-object-storage/admin/content/installing-and-configuring-the-proxy-node.html | 00:18 |
| maplebed | and http://docs.openstack.org/api/openstack-object-storage/1.0/content/authentication-object-dev-guide.html for the dev stuff. | 00:18 |
| maplebed | notmyname: my coworker was reporting that using the account:user and pass didn't work with the x-auth headers; let me try them to confirm.. back in 3. | 00:19 |
| notmyname | maplebed: the x-auth* vs x-storage* is, as AaronSchulz said, is for legacy clients (since swift started life as a product at rackspace) | 00:20 |
| notmyname | maplebed: use x-auth-[user|key] | 00:20 |
| notmyname | and x-auth-token | 00:20 |
| maplebed | ok, so the simple answer is that x-storage-user and x-storage-pass have been renamed to x-auth-user and x-auth-key. | 00:21 |
| maplebed | also, my test with curl passed. | 00:21 |
| maplebed | (passed => I get back an x-storage-url and x-storage-token) | 00:21 |
| maplebed | it's interesting that I get back an x-storage-token in addition to x-auth-token (they're the same); the api docs only mention x-auth-token. | 00:22 |
| maplebed | notmyname: thanks; I think I am unconfused for the time being. | 00:22 |
| *** MarkA2 has quit IRC | 00:22 | |
| notmyname | maplebed: http://paste.openstack.org/show/3594/ | 00:23 |
| *** jakedahn has quit IRC | 00:23 | |
| notmyname | maplebed: ya, x-auth-token and x-storage-token have the same relationship as the other ones. use x-auth-token | 00:24 |
| maplebed | that paste is very similar to what I have. , though the end of the config URL is AUTH_system rather than AUTH_abc | 00:25 |
| maplebed | what is the stuff after the underscore? | 00:25 |
| *** lloydde has quit IRC | 00:25 | |
| *** lloydde has joined #openstack | 00:26 | |
| uvirtbot | New bug: #898867 in horizon "Launch instance not working after launch instance -> cancel." [Undecided,New] https://launchpad.net/bugs/898867 | 00:26 |
| *** adjohn has joined #openstack | 00:26 | |
| *** pixelbeat has joined #openstack | 00:27 | |
| *** AaronSchulz has quit IRC | 00:29 | |
| *** AaronSchulz has joined #openstack | 00:30 | |
| *** jakedahn has joined #openstack | 00:30 | |
| *** jakedahn has quit IRC | 00:31 | |
| *** tdi has joined #openstack | 00:31 | |
| *** tdi_ has quit IRC | 00:32 | |
| *** rnorwood has quit IRC | 00:33 | |
| *** mattray has joined #openstack | 00:33 | |
| notmyname | maplebed: you mean the "abc" in my example | 00:33 |
| notmyname | ? | 00:33 |
| maplebed | yeah. | 00:33 |
| *** lloydde has quit IRC | 00:35 | |
| *** lloydde has joined #openstack | 00:35 | |
| notmyname | maplebed: a swift account string is made up of 2 parts, separated by the first underscore. the first part ("AUTH" in my example) is the reseller. this allows you to have multiple auth systems for distinct accounts in the cluster without having to ask every auth system about every request. the second part ("abc" in my example) is the unique account string. there's nothing special about the second part. normally we use uuid.uuid4().hex to generate it | 00:36 |
| maplebed | I thought the account part was the testinga portion of the string. | 00:38 |
| maplebed | (so the auth token was account:user) | 00:38 |
| maplebed | s/auth token/x-auth-user header/ | 00:39 |
| maplebed | I saw this term 'reseller' in some docs as well. Can you explain (or point me to docs that explain) what it means? | 00:40 |
| pweiss | anyone know if devstack is broken? mine is stuck: + sudo PIP_DOWNLOAD_CACHE=/var/cache/pip pip install --use-mirrors Django==1.3 django-mailer django-nose==0.1.2 django-nose-selenium django-registration==0.7 pycrypto==2.3 sqlalchemy-migrate python-cloudfiles -e git+https://github.com/cloudbuilders/openstackx.git#egg=openstackx -e git+https://github.com/jacobian/openstack.compute.git#egg=openstack -e git+https://github.com/4P/pyt | 00:42 |
| maplebed | (I think I understand from your comment that it means you can have something like AUTHONE, AUTHTWO, AUTHTHREE and they'll have distinct users | 00:42 |
| *** tungvs has joined #openstack | 00:45 | |
| *** pixelbeat has quit IRC | 00:46 | |
| *** whit has joined #openstack | 00:49 | |
| *** adjohn has quit IRC | 00:50 | |
| *** cdub has quit IRC | 00:51 | |
| *** adjohn has joined #openstack | 00:53 | |
| *** dragondm has quit IRC | 00:54 | |
| *** adjohn has quit IRC | 00:55 | |
| *** jaypipes has joined #openstack | 00:56 | |
| *** rustam has joined #openstack | 00:56 | |
| *** MarcMorata has quit IRC | 00:57 | |
| *** AlanClark has quit IRC | 00:58 | |
| *** dotdevops has quit IRC | 01:02 | |
| *** yshh has joined #openstack | 01:03 | |
| *** pringles_ has quit IRC | 01:04 | |
| *** pringles has joined #openstack | 01:04 | |
| *** longhudou has joined #openstack | 01:06 | |
| *** adjohn has joined #openstack | 01:07 | |
| *** adjohn has quit IRC | 01:07 | |
| notmyname | maplebed: the account is testinga:testingb | 01:07 |
| notmyname | maplebed: the auth token is what the auth server returns and what you pass for subsequent requests | 01:07 |
| notmyname | maplebed: ya, you are right about the resellers | 01:08 |
| maplebed | this was the part that confused me: "the second part ("abc" in my example) is the unique account string." | 01:08 |
| *** livemoon has joined #openstack | 01:08 | |
| notmyname | maplebed: while it's possible to have multiple auth servers without resellers, having resellers allows some efficiencies | 01:08 |
| *** yshh has quit IRC | 01:09 | |
| *** clopez has joined #openstack | 01:09 | |
| *** yshh has joined #openstack | 01:09 | |
| *** woleium has quit IRC | 01:09 | |
| notmyname | maplebed: for example, at rackspace we have a reseller for the rackspace cloud users and a reseller for jungle disk | 01:09 |
| *** microtosh has quit IRC | 01:09 | |
| maplebed | makes sense. | 01:09 |
| notmyname | maplebed: the siwft account string is RESELLER_UNIQUEID | 01:10 |
| maplebed | ah, so it's a human-readable vs. internal-unique-id thing? | 01:10 |
| maplebed | I'll use testinga:testingb and internally that turns into AUTH_abc | 01:10 |
| *** jog0 has joined #openstack | 01:11 | |
| notmyname | correct. it's the difference in the auth account and the swift cluster account | 01:13 |
| notmyname | one isn't "human readable" compared to the other, they are just different | 01:13 |
| notmyname | the auth server ties them together | 01:13 |
| maplebed | gotcha. | 01:13 |
| maplebed | I was missing the auth account vs. swift account distinction. | 01:14 |
| maplebed | Thanks a bunch; your descriptions have been very helpful. | 01:14 |
| *** vladimir3p has quit IRC | 01:15 | |
| *** pradeep1 has joined #openstack | 01:17 | |
| longhudou | I can ping the instance running on the controller node from the controller node ,while from the compute node I can't ping | 01:17 |
| longhudou | has anybody run into this circumstance before? | 01:17 |
| *** wilmoore has quit IRC | 01:17 | |
| *** adjohn has joined #openstack | 01:21 | |
| *** adjohn has quit IRC | 01:22 | |
| *** pweiss has quit IRC | 01:26 | |
| *** huslage has joined #openstack | 01:27 | |
| *** adjohn has joined #openstack | 01:28 | |
| *** stuntmachine has joined #openstack | 01:28 | |
| *** adjohn has quit IRC | 01:28 | |
| *** mattray has quit IRC | 01:28 | |
| *** fridim_ has quit IRC | 01:31 | |
| *** dholmberg has quit IRC | 01:34 | |
| *** adjohn has joined #openstack | 01:38 | |
| *** rods has quit IRC | 01:38 | |
| *** bengrue has quit IRC | 01:40 | |
| *** woleium has joined #openstack | 01:41 | |
| *** llang629 has quit IRC | 01:41 | |
| *** Kiall has quit IRC | 01:46 | |
| tungvs | longhudou: which networkmanager are you using ? | 01:46 |
| longhudou | flatdhcp | 01:47 |
| tungvs | what kind of ipaddress of the vmguest are you using to ping ? | 01:47 |
| tungvs | fixed or floating ? | 01:48 |
| longhudou | fixed | 01:48 |
| longhudou | 10.0.0.0/24 | 01:48 |
| tungvs | what's compute-node IP ? | 01:48 |
| longhudou | 192.168.1.13 | 01:49 |
| *** amdtech has left #openstack | 01:49 | |
| tungvs | Has compute-node got any 10.0.0.0/24 IP ? | 01:50 |
| *** _adjohn has joined #openstack | 01:50 | |
| *** blakeyeager has quit IRC | 01:51 | |
| longhudou | what do you mean by that ,I am confused ,10.0.0.0/24 is a network that instance get ip from? isn't that? | 01:51 |
| longhudou | can compute-node got an 10.0.0.0/24 IP? | 01:51 |
| longhudou | how to ? | 01:51 |
| *** Ryan_Lane has quit IRC | 01:52 | |
| *** adjohn has quit IRC | 01:52 | |
| *** _adjohn is now known as adjohn | 01:52 | |
| tungvs | in my implement, I've assigned my bridge 1 IP in the fixed-range so the host can communicate with the vmguest | 01:52 |
| *** Ryan_Lane has joined #openstack | 01:52 | |
| *** MarkA2 has joined #openstack | 01:54 | |
| *** songyan has joined #openstack | 01:54 | |
| longhudou | we don't use public IP, | 01:54 |
| tungvs | fixed-rage should be private IP, not public | 01:55 |
| *** CaptTofu has quit IRC | 01:55 | |
| *** po has quit IRC | 01:55 | |
| longhudou | I know that,we are in a LAN,and our IP are 192,we using NAT to get to the public,and I create the fixed IP range with 10.0.0.0 | 01:56 |
| longhudou | do I need to create the fixed IP with 192? | 01:57 |
| tungvs | in your case, I suggest check to see the compute-host can ping any IP in the 10.0.0.0/24, if yes, then dig into the controller iptables to see if it block the compute-host IP, if iptables is OK, then you should use tcpdump to check what happend | 01:57 |
| *** woleium has quit IRC | 01:57 | |
| *** jog0 has quit IRC | 01:57 | |
| *** woleium has joined #openstack | 01:57 | |
| longhudou | from the compute-node i can not ping even 10.0.0.1 | 01:58 |
| tungvs | show me your "ip addr" in the controller and the compute-node | 01:59 |
| longhudou | controller 192.168.1.91,compute 192.168.1.13 | 02:00 |
| tungvs | no, type "ip addr" in your terminal and show us the result | 02:00 |
| longhudou | 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet 169.254.169.254/32 scope link lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1 000 lin | 02:01 |
| longhudou | controller | 02:01 |
| longhudou | 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet 169.254.169.254/32 scope link lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP ql en 1000 | 02:02 |
| longhudou | compute | 02:02 |
| tungvs | please use paste.bin | 02:03 |
| longhudou | I am sorry to say that I don't know paste.bin | 02:04 |
| longhudou | where can I find it | 02:05 |
| longhudou | or how to use it | 02:05 |
| maplebed | paste.openstack.org | 02:05 |
| tungvs | http://pastebin.com | 02:05 |
| maplebed | or pastie.org or ... ;) (there are many around the net) | 02:05 |
| longhudou | http://paste.openstack.org/show/3595/ | 02:07 |
| longhudou | can you find that? | 02:07 |
| tungvs | longhudou: you should assign an 10.0.0.0/24 IP, like 10.0.0.100 to your compute bridge | 02:08 |
| *** krow has joined #openstack | 02:09 | |
| tungvs | and check whether compute br0 is contain eth0 or not | 02:09 |
| longhudou | how to assign an 10.0.0.0/24 IP to compute bridge,I am just a fresh bird | 02:10 |
| longhudou | :< | 02:10 |
| tungvs | ok, ip addr add 10.0.0.100/24 dev br0 | 02:11 |
| tungvs | and show us "brctl show" | 02:11 |
| longhudou | http://paste.openstack.org/show/3596/ | 02:13 |
| tungvs | and "brctl show" of the controller also | 02:14 |
| tungvs | it's short, you could paste it here directly | 02:14 |
| *** jakedahn has joined #openstack | 02:14 | |
| longhudou | br0 8000.f46d0475fd80 no eth0 vnet0 virbr0 8000.000000000000 yes | 02:15 |
| longhudou | this is the controller | 02:15 |
| tungvs | ok then | 02:15 |
| tungvs | try to ping again | 02:15 |
| *** maplebed has quit IRC | 02:16 | |
| longhudou | Okļ¼this time I can ping the instance running on the controller node from the compute node | 02:17 |
| longhudou | thanks a lot | 02:17 |
| longhudou | but there is another problem,@tungvs | 02:18 |
| longhudou | on the controller node I can see an instance running on the compute node ,But I cannot ping | 02:18 |
| *** jakedahn has quit IRC | 02:18 | |
| tungvs | just use euca-get-console | 02:20 |
| *** mwhooker has quit IRC | 02:20 | |
| tungvs | to check if the vmguest has booted up properly or not | 02:20 |
| *** huslage has quit IRC | 02:20 | |
| longhudou | is shows:::::2011-12-02T02:20:55Z | 02:21 |
| tungvs | anything else ? | 02:21 |
| longhudou | no | 02:21 |
| longhudou | only this | 02:21 |
| tungvs | the it cannot boot up | 02:21 |
| longhudou | it's wrong? | 02:21 |
| *** AaronSchulz has quit IRC | 02:22 | |
| *** clopez has quit IRC | 02:22 | |
| tungvs | what's your compute-node arch, or x64 ? | 02:22 |
| tungvs | I mean 32 or 64 bit ? | 02:22 |
| *** cp16net has quit IRC | 02:22 | |
| longhudou | I install ubuntu11.04 32bit | 02:22 |
| longhudou | do you mean that? or the hardware? | 02:23 |
| tungvs | and I guess you're using 64 bit image to create vmguest ? | 02:23 |
| longhudou | oh,no,the host system and the vmguest is the same | 02:24 |
| tungvs | what hypervisor are you using ? | 02:25 |
| longhudou | kvm | 02:25 |
| *** spiffxp has quit IRC | 02:25 | |
| tungvs | are your hardware support that ? | 02:25 |
| longhudou | yes | 02:25 |
| *** bencherian has quit IRC | 02:25 | |
| longhudou | maybe it is caused by the wrong image? | 02:26 |
| tungvs | do a "ps -ef | grep kvm" in your compute-node | 02:26 |
| longhudou | I create it manually | 02:26 |
| tungvs | can you see any process ? | 02:26 |
| *** bencherian has joined #openstack | 02:26 | |
| longhudou | root 1025 2 0 Nov30 ? 00:00:00 [kvm-irqfd-clean] | 02:26 |
| tungvs | anything else ? | 02:26 |
| *** pradeep1 has quit IRC | 02:26 | |
| longhudou | 116 4132 1 0 09:24 ? 00:00:36 /usr/bin/kvm -S -M pc-0.14 -enable-kvm | 02:27 |
| longhudou | root 4133 2 0 09:24 ? 00:00:00 [kvm-pit-wq] | 02:27 |
| *** jakedahn has joined #openstack | 02:27 | |
| *** rustam has quit IRC | 02:27 | |
| tungvs | just paste it all | 02:27 |
| *** bencherian has quit IRC | 02:28 | |
| longhudou | http://paste.openstack.org/show/3600/ | 02:28 |
| longhudou | Ok, it is on the page3600 | 02:29 |
| tungvs | use VNC to connect to your vmguest | 02:29 |
| tungvs | maybe something's worng with your image | 02:29 |
| tungvs | you should download an ec2 image to create vmguest first | 02:30 |
| tungvs | to make sure your nova setup is ok | 02:30 |
| longhudou | i think something is wrong with my image,too | 02:31 |
| *** reed has quit IRC | 02:32 | |
| *** andrewbogott has quit IRC | 02:32 | |
| *** stuntmachine has quit IRC | 02:32 | |
| *** Kiall has joined #openstack | 02:34 | |
| longhudou | I create it manually ,and in the installation process, I create more than one partition(root,swap,for example),and after that, i did not adjust the image,does openstack only support ext4 image? | 02:34 |
| longhudou | @tungvs | 02:34 |
| larissa | longhudou: Error: "tungvs" is not a valid command. | 02:34 |
| *** redconnection has joined #openstack | 02:34 | |
| *** misha has joined #openstack | 02:34 | |
| *** misha is now known as Guest87127 | 02:34 | |
| longhudou | tungvs | 02:35 |
| *** misheska has quit IRC | 02:35 | |
| *** clauden_ has quit IRC | 02:37 | |
| tungvs | I suggest you stick with the guide if you don't know what you're doing. I think you have to adjust the image because nova will mount the image as root partition, so the image should be read as a partition, hence must be adjusted | 02:37 |
| tungvs | can anyone please verify that for us ? | 02:38 |
| *** stuntmachine has joined #openstack | 02:38 | |
| longhudou | Ok, then i will download ec2 image to have a try | 02:38 |
| *** cmagina has quit IRC | 02:38 | |
| longhudou | thanks very much | 02:38 |
| *** camm has quit IRC | 02:39 | |
| *** clauden_ has joined #openstack | 02:40 | |
| *** stuntmachine has quit IRC | 02:40 | |
| *** Guest87127 has quit IRC | 02:44 | |
| *** osier has joined #openstack | 02:52 | |
| *** adjohn has quit IRC | 02:58 | |
| *** AaronSchulz has joined #openstack | 03:00 | |
| *** camm has joined #openstack | 03:02 | |
| *** AaronSchulz has quit IRC | 03:04 | |
| *** agrajag has quit IRC | 03:08 | |
| *** agrajag has joined #openstack | 03:09 | |
| *** packet__ has quit IRC | 03:15 | |
| *** andrewbogott has joined #openstack | 03:18 | |
| *** janpy has joined #openstack | 03:22 | |
| *** redconnection has quit IRC | 03:26 | |
| *** misheska has joined #openstack | 03:28 | |
| *** misheska has quit IRC | 03:28 | |
| *** andrewbogott has quit IRC | 03:32 | |
| *** hadrian has quit IRC | 03:32 | |
| *** rnorwood has joined #openstack | 03:38 | |
| *** dolphm has joined #openstack | 03:40 | |
| *** cmagina has joined #openstack | 03:42 | |
| *** wilmoore has joined #openstack | 03:43 | |
| *** agrajag has quit IRC | 03:45 | |
| *** agrajag has joined #openstack | 03:45 | |
| *** koolhead17 has quit IRC | 03:45 | |
| uvirtbot | New bug: #898935 in keystone "bin/keystone doesn't respect port settings" [Undecided,In progress] https://launchpad.net/bugs/898935 | 03:46 |
| *** rnorwood has left #openstack | 03:46 | |
| *** bombay has joined #openstack | 03:48 | |
| *** Ryan_Lane has quit IRC | 03:49 | |
| *** stuntmachine has joined #openstack | 03:52 | |
| *** rsampaio has joined #openstack | 03:52 | |
| *** robbiew has quit IRC | 04:03 | |
| *** jdurgin has quit IRC | 04:03 | |
| *** misheska has joined #openstack | 04:03 | |
| *** misheska has quit IRC | 04:04 | |
| *** janpy has quit IRC | 04:04 | |
| *** janpy has joined #openstack | 04:05 | |
| tungvs | I've created an instance_type with swap space about 512 MB. But when I boot vmguest up and do "fdisk -l" as root, it shows /dev/vda,vdb,vdc dont contain valid partition table. And "swapon -s" show nothing, means no swap. Can anyone explain this ? | 04:10 |
| *** stuntmachine has quit IRC | 04:11 | |
| *** dolphm has quit IRC | 04:14 | |
| *** CaptTofu has joined #openstack | 04:15 | |
| *** packet_ has quit IRC | 04:15 | |
| *** jj0hns0n_ has quit IRC | 04:19 | |
| *** ldlework has joined #openstack | 04:19 | |
| *** jj0hns0n has joined #openstack | 04:20 | |
| *** scottjg has joined #openstack | 04:21 | |
| *** lborda has joined #openstack | 04:21 | |
| *** jj0hns0n has quit IRC | 04:22 | |
| *** vidd has joined #openstack | 04:33 | |
| *** Rajaram has joined #openstack | 04:33 | |
| *** vidd has quit IRC | 04:34 | |
| *** vidd has joined #openstack | 04:34 | |
| *** jfluhmann has joined #openstack | 04:38 | |
| *** jakedahn has quit IRC | 04:39 | |
| *** nati2 has quit IRC | 04:39 | |
| *** janpy has quit IRC | 04:43 | |
| *** vidd-away has joined #openstack | 04:44 | |
| *** vidd-away has joined #openstack | 04:44 | |
| *** vidd has quit IRC | 04:45 | |
| *** jakedahn has joined #openstack | 04:51 | |
| *** Ryan_Lane has joined #openstack | 04:51 | |
| *** jakedahn has quit IRC | 04:51 | |
| *** supriya has joined #openstack | 04:53 | |
| *** adjohn has joined #openstack | 04:54 | |
| *** adjohn has quit IRC | 04:59 | |
| *** Ryan_Lane is now known as Ryan_Lane|away | 05:01 | |
| *** lborda has quit IRC | 05:05 | |
| *** PeteDaGuru has quit IRC | 05:15 | |
| *** ldlework has quit IRC | 05:19 | |
| jasona | hmm | 05:19 |
| *** koolhead17 has joined #openstack | 05:22 | |
| *** pradeep has joined #openstack | 05:26 | |
| *** koolhead17 has quit IRC | 05:31 | |
| *** wilmoore has quit IRC | 05:32 | |
| *** pradeep has quit IRC | 05:34 | |
| *** dpippenger has quit IRC | 05:34 | |
| *** tungvs has quit IRC | 05:35 | |
| *** tungvs has joined #openstack | 05:42 | |
| *** mjfork has quit IRC | 05:44 | |
| *** rsampaio has quit IRC | 05:46 | |
| *** rsampaio has joined #openstack | 05:49 | |
| *** pradeep1 has joined #openstack | 05:49 | |
| *** pradeep1 has joined #openstack | 05:50 | |
| *** ghanima has quit IRC | 05:56 | |
| *** jward has joined #openstack | 05:56 | |
| *** tmichael has quit IRC | 05:59 | |
| *** jward has joined #openstack | 05:59 | |
| *** tmichael has joined #openstack | 06:00 | |
| *** Frodent has quit IRC | 06:06 | |
| *** janpy has joined #openstack | 06:06 | |
| *** reed has joined #openstack | 06:08 | |
| *** rsampaio has quit IRC | 06:10 | |
| *** rsampaio has joined #openstack | 06:12 | |
| *** jward has quit IRC | 06:13 | |
| *** supriya has quit IRC | 06:14 | |
| *** fulcrum has joined #openstack | 06:16 | |
| *** jward has joined #openstack | 06:17 | |
| *** supriya has joined #openstack | 06:18 | |
| *** reed has quit IRC | 06:19 | |
| *** nerens has joined #openstack | 06:24 | |
| *** zefry has joined #openstack | 06:26 | |
| zefry | hello all | 06:26 |
| zefry | can you help me about cloud-publish-image i386 serverfinal.img bucket1 | 06:27 |
| zefry | the result is : failed to check for existing manifest | 06:28 |
| zefry | anyone, can you help me? | 06:28 |
| zefry | :( | 06:28 |
| *** arBmind has joined #openstack | 06:33 | |
| *** tungvs has quit IRC | 06:33 | |
| *** tungvs has joined #openstack | 06:33 | |
| *** fulcrum has quit IRC | 06:35 | |
| *** objectiveous has joined #openstack | 06:36 | |
| *** rsampaio has quit IRC | 06:40 | |
| zefry | anyone, can you help me? | 06:51 |
| tungvs | zefry: I guess you should check if the glance image service runs or not, the config in nova.conf about glance IP:port, and the sql databse which glance connects to | 06:56 |
| *** mindpixel has joined #openstack | 07:07 | |
| *** wilmoore has joined #openstack | 07:10 | |
| *** pradeep1 has quit IRC | 07:16 | |
| *** jakedahn has joined #openstack | 07:20 | |
| *** jakedahn_ has joined #openstack | 07:22 | |
| *** jakedahn has quit IRC | 07:25 | |
| *** jakedahn_ is now known as jakedahn | 07:25 | |
| *** wilmoore has quit IRC | 07:25 | |
| *** jakedahn has quit IRC | 07:28 | |
| *** pradeep1 has joined #openstack | 07:32 | |
| *** javiF has joined #openstack | 07:33 | |
| *** matiu has joined #openstack | 07:34 | |
| tungvs | I've created an instance_type with swap space about 512 MB. But when I boot vmguest up, run "swapon -s" show nothing, means no swap. Can anyone explain this ? | 07:36 |
| *** mgoldmann has joined #openstack | 07:38 | |
| *** woleium has quit IRC | 07:44 | |
| tungvs | anyone knows what's in the /var/lib/nova/instances/_base ? I find its size is huge | 07:47 |
| *** tmichael has quit IRC | 07:47 | |
| *** tmichael has joined #openstack | 07:48 | |
| *** jedi4ever has joined #openstack | 07:49 | |
| *** lloydde has quit IRC | 07:49 | |
| *** miclorb_ has quit IRC | 07:50 | |
| *** kaigan_ has joined #openstack | 07:50 | |
| *** marcuz has joined #openstack | 07:51 | |
| marcuz | good morning | 07:51 |
| uvirtbot | New bug: #898983 in horizon "cancel button on launch instance window really hard to see" [Undecided,New] https://launchpad.net/bugs/898983 | 07:51 |
| *** arBmind has quit IRC | 07:54 | |
| *** adiantum has joined #openstack | 07:54 | |
| *** pixelbeat has joined #openstack | 07:57 | |
| *** foexle has joined #openstack | 07:58 | |
| *** adiantum has quit IRC | 08:01 | |
| foexle | ahoi @all | 08:01 |
| *** GheRivero has quit IRC | 08:03 | |
| tungvs | My nova-compute sometimes just hangup. When I make nova-compute restart, it hangs at the log line "Connecting to libvirt: qemu:///system.." | 08:04 |
| tungvs | Any idea ? | 08:04 |
| *** GheRivero has joined #openstack | 08:06 | |
| *** reidrac has joined #openstack | 08:07 | |
| *** TheOsprey has joined #openstack | 08:07 | |
| *** binbash_ has quit IRC | 08:11 | |
| *** binbash_ has joined #openstack | 08:12 | |
| *** krow has quit IRC | 08:13 | |
| *** sloop has quit IRC | 08:14 | |
| *** Ramonster has joined #openstack | 08:14 | |
| *** sloop has joined #openstack | 08:15 | |
| *** sloop has joined #openstack | 08:15 | |
| *** jfluhmann has quit IRC | 08:18 | |
| *** sloop has quit IRC | 08:24 | |
| *** sloop has joined #openstack | 08:26 | |
| *** javiF has quit IRC | 08:27 | |
| *** guigui has joined #openstack | 08:32 | |
| *** sloop has quit IRC | 08:39 | |
| *** popux has joined #openstack | 08:47 | |
| *** vdo has joined #openstack | 08:51 | |
| *** apurva has joined #openstack | 08:52 | |
| *** oubiwann has quit IRC | 08:52 | |
| *** sloop has joined #openstack | 08:52 | |
| apurva | need help getting invalid literal for int() with base 10: '' error while running euca-authorize -P icmp -t -1:-1 default command | 08:54 |
| *** Razique has joined #openstack | 08:54 | |
| *** uksysadmin has joined #openstack | 08:56 | |
| *** janpy has quit IRC | 08:58 | |
| *** dobber has joined #openstack | 08:59 | |
| *** PotHix has joined #openstack | 09:08 | |
| Sander^work | I found an bug in the docs.. condradiction theese two places: http://keystone.openstack.org/configuring.html and http://docs.openstack.org/diablo/openstack-identity/admin/content/creating-tenants-users-roles-tokens-and-endpoints.html | 09:25 |
| Sander^work | in the endpointTemplates i'm unsure if I should include the tentant_id at the end of the url for nova | 09:25 |
| Sander^work | One of the docs say one thing, the other say something else... Anyone know what's right? | 09:26 |
| *** nacx has joined #openstack | 09:28 | |
| *** PiotrSikora has quit IRC | 09:30 | |
| apurva | any one can please help me on euca-authorize command error | 09:31 |
| *** PiotrSikora has joined #openstack | 09:32 | |
| Sander^work | Anyone around that have set up keystone? | 09:36 |
| *** scottjg has quit IRC | 09:38 | |
| marcuz | Sander^work: agree, it generates an exception on line 9 of "keystone/manage/api.py", isn't it? | 09:40 |
| zefry | hello @all | 09:43 |
| *** fulcrum has joined #openstack | 09:44 | |
| zefry | I have trouble with uploading to openstack | 09:44 |
| Sander^work | marcuz, I don't know where to look for that exception.. in which log files? | 09:44 |
| *** songyan has quit IRC | 09:45 | |
| Sander^work | marcuz, do you know what the right value of nova endpointTemplates is? | 09:45 |
| zefry | command is : cloud-publish-image i386 serverfinal.img bucket1 | 09:45 |
| zefry | and result is: failed to check for existing manifest | 09:46 |
| zefry | do you have a solutions for it? | 09:46 |
| marcuz | Sander^work: can't go in detail, i was working on it a couple of days ago and i figured out it's breaking when creating the user, i misread your question, sorry :) | 09:46 |
| zefry | marcuz: can you help me?? :( | 09:47 |
| Sander^work | marcuz, I'm unable to log in to the openstack dashboard.. just gives me same page back. | 09:48 |
| *** rustam has joined #openstack | 09:57 | |
| *** apurva has quit IRC | 10:04 | |
| *** lzyeval has quit IRC | 10:07 | |
| *** katkee has joined #openstack | 10:08 | |
| *** yshh has quit IRC | 10:10 | |
| tungvs | zefry: I guess you should check if the glance image service runs or not, the config in nova.conf about glance IP:port, and the sql databse which glance connects to | 10:11 |
| *** clopez has joined #openstack | 10:13 | |
| zefry | tungvs: glance is running, | 10:14 |
| Razique | zefry: it's a custom image ? | 10:14 |
| *** adiantum has joined #openstack | 10:15 | |
| uvirtbot | New bug: #899056 in openstack-ci "python-novaclient needs milestone-proposed Jenkins jobs" [High,Confirmed] https://launchpad.net/bugs/899056 | 10:16 |
| zefry | Razique: I don't understand about custom image.. | 10:17 |
| Razique | zefry:where did u get the image from ? | 10:18 |
| zefry | from release.ubuntu.com | 10:18 |
| *** mcclurmc has quit IRC | 10:18 | |
| Razique | ok that's the issue here | 10:19 |
| Razique | mmm wait | 10:19 |
| zefry | Razique: okey, thank you before :) | 10:19 |
| *** mcclurmc has joined #openstack | 10:20 | |
| Razique | zefry: do you have a man for it ? | 10:20 |
| Razique | I can't find it | 10:20 |
| *** BasTichelaar has joined #openstack | 10:21 | |
| Razique | in fact I think it looks for "cloudified" images | 10:21 |
| Razique | eg. splitted kernel, ramdisk and image | 10:21 |
| *** livemoon has quit IRC | 10:21 | |
| Razique | since you are trying to upload an ISO (non "cloudified" image :D ) the command can't find the manifest (xml file that describes the structure of your image | 10:22 |
| Razique | you would want to use these images : | 10:22 |
| Razique | http://cloud-images.ubuntu.com/server/ | 10:22 |
| Razique | that will work for sure | 10:22 |
| zefry | okey I'm try .. | 10:22 |
| foexle | heyho Razique how are you :) | 10:23 |
| foexle | ? | 10:23 |
| Razique | foexle: :D | 10:23 |
| Razique | hi dude | 10:23 |
| Razique | we missed each other 2 days ago! | 10:23 |
| Razique | I was up 'til 2 am | 10:23 |
| zefry | Razique: thanks for sharing :) | 10:24 |
| foexle | yeah i know :( .... i sleepd 30 mins before youe came on :/ | 10:24 |
| Razique | hehe tonight then ? | 10:24 |
| Razique | zefry: sure | 10:24 |
| *** javiF has joined #openstack | 10:24 | |
| foexle | Razique: oh today i'll go out with friends | 10:25 |
| foexle | Razique: because my friends says every day ... "Hey heiko where are you, why you don't have time" and so on ;) | 10:25 |
| *** rods has joined #openstack | 10:26 | |
| BasTichelaar | hey guys | 10:26 |
| Razique | hehe yah | 10:26 |
| foexle | Razique: so should we do next week .... if you have time just say me ... EY heiko lets go hahahahaha | 10:26 |
| BasTichelaar | anyone has setup a multi-zone cluster? | 10:26 |
| Razique | we geek don't use "Social Life" as PPA often | 10:26 |
| foexle | ^^ | 10:26 |
| uvirtbot | foexle: Error: "^" is not a valid command. | 10:26 |
| zefry | Razique: what is a good format to run? tar.gz or .img ?? | 10:26 |
| Razique | hey BasTichelaar :) | 10:26 |
| Razique | zefry: tar.gz for cloud-utils | 10:27 |
| BasTichelaar | Razique: hey, how are you doing? | 10:27 |
| zefry | Razique: thanks :) | 10:27 |
| Razique | BasTichelaar: playing with swap provisionning for instances | 10:27 |
| Razique | :p | 10:27 |
| Razique | Swapless instances are scary | 10:27 |
| BasTichelaar | Razique: swap provisioning? whats that? | 10:27 |
| BasTichelaar | Razique: ah ok | 10:27 |
| Razique | so I'm working on a way to do swap provisionning based on the actual RAM | 10:27 |
| Razique | nothing much but I've something that works here | 10:28 |
| BasTichelaar | Razique: shouldnt be to difficult, are you gonna hack it into openstack? | 10:28 |
| Razique | https://github.com/Razique/BashStuff/blob/master/SCR_5016_V00_NUAC-OPENSTACK-Swap-provisionning.sh | 10:28 |
| Razique | BasTichelaar: no, I don't know Python much | 10:28 |
| Razique | so I created a bash script meant to be run from rc.local atm | 10:28 |
| BasTichelaar | Razique: its quite simple, there is a file disk.py where the commands are run for a created instance | 10:29 |
| BasTichelaar | you can modify them to include a swap partition | 10:29 |
| Razique | oh great! | 10:29 |
| BasTichelaar | its just cfdisk commands | 10:29 |
| Razique | thanks i'll look for that | 10:29 |
| Razique | what about you ? still working on multi-zones ? | 10:29 |
| BasTichelaar | Razique: yes, but I have a very simple use case, that doesnt seem to be implemented | 10:30 |
| BasTichelaar | I have two zones, one in amsterdam and one in zurich | 10:30 |
| Razique | :/ | 10:30 |
| Razique | bummer | 10:30 |
| BasTichelaar | and I wanna be able to specify the zone where an instance should live | 10:30 |
| BasTichelaar | but it seems I just have to connect to a different API if I'm correct | 10:30 |
| BasTichelaar | but then projects are spread across multiple zones, but there is no global knowledge | 10:31 |
| Razique | yup | 10:31 |
| Razique | yah i think you are rigth BasTichelaar | 10:31 |
| BasTichelaar | anyway, I have opened a question for it | 10:31 |
| BasTichelaar | hope sandy will answer | 10:31 |
| *** zefry has quit IRC | 10:32 | |
| coli | bastichelaar: look at multiclusterzones | 10:32 |
| Razique | I'm also waiting for instance's resizing under KVM | 10:32 |
| Razique | that would be cool | 10:32 |
| coli | you need two zones as child of parent zone | 10:32 |
| coli | then talk to api for the parent zone | 10:32 |
| BasTichelaar | coli: yes, but you cannot specify the zone manually right? | 10:32 |
| BasTichelaar | coli: I did implement it, but found no way to do what I want | 10:33 |
| BasTichelaar | coli: or should I fire up instances in child zones, and for the overview connect to the parent zone? | 10:33 |
| coli | bastichelaar: hmm... look into api code and see if there is a zone parameter for create (DON't look into documentation which is way outdated) | 10:34 |
| tungvs | Razique: talking about swap in instance, I've created an instance_type including swap. When I boot vmguest up, I see an extra /dev/vdb, but it contains no swap partition, can you explain the proper way to have swap in vmguest ? | 10:34 |
| BasTichelaar | coli: well, there is one for availability zones in python-novaclient | 10:35 |
| BasTichelaar | but not for the multi-cluster zones | 10:35 |
| coli | bastichelaar: I would expect one to be there, otherwise it would be no point for zones except only for internal usage. | 10:35 |
| coli | on the other hand, what you are saying that talking to two api's is not that bad | 10:35 |
| coli | do yo uuse dashboard or your own frontend ? | 10:36 |
| BasTichelaar | coli: no, thats what I thought | 10:36 |
| BasTichelaar | coli: we're not using a frontend, but the python-novaclient api | 10:36 |
| coli | bastichelarr: I will be playing with it this weekend if I find time :-) | 10:36 |
| BasTichelaar | coli: cool :) | 10:36 |
| coli | so look into nova-api code as well and see if there is zone parameter for create and if it's used later on :-) | 10:37 |
| BasTichelaar | https://answers.launchpad.net/nova/+question/180662 | 10:37 |
| BasTichelaar | yes, I'm now reading through the api code | 10:37 |
| Razique | tungvs: yah... it only provides a disk for swap | 10:38 |
| Razique | but doesn't enable it | 10:38 |
| Razique | so format it and swapon-it | 10:39 |
| tungvs | Razique: sure I will, tks | 10:39 |
| coli | anybody used RADOS (rbd) as storage for nova ? | 10:39 |
| Razique | that's also why my swap creation is scripted by now | 10:39 |
| tungvs | Razique: script ? How can you script to do something inside a newly created instance ? I'm interested to know. | 10:41 |
| Razique | rc.local | 10:43 |
| Razique | I put a ref to my swap script on /Etc/rc.local | 10:43 |
| Razique | on boot it is executed | 10:43 |
| Razique | as the metadatas are retrieved | 10:43 |
| *** livemoon has joined #openstack | 10:44 | |
| BasTichelaar | Razique: idea: provide the size of the swap in the metadata :) | 10:44 |
| livemoon | hi,all | 10:44 |
| Razique | hi livemoon :) | 10:44 |
| livemoon | hi,Razique | 10:44 |
| Razique | BasTichelaar: how ? | 10:44 |
| Razique | what's up man ? | 10:44 |
| foexle | hi livemoon | 10:44 |
| BasTichelaar | Razique: by setting the metada and reading them out (they get written to /meta.js inside the instance) | 10:45 |
| livemoon | hi,foexle | 10:45 |
| Razique | BasTichelaar: what would be the gain ? | 10:45 |
| Razique | more of a "structural" better way ? | 10:45 |
| Razique | so swap should be seen as a metadata also ? | 10:45 |
| BasTichelaar | Razique: well, then you can at least specify the size of the swap on instance creation | 10:46 |
| BasTichelaar | Razique: dont know if its needed for your usecase | 10:46 |
| Razique | BasTichelaar: yah, my script does that too | 10:46 |
| Razique | but I'm goan update my base images | 10:46 |
| Razique | so they include the script and a first swapfile | 10:47 |
| livemoon | your new script about swap? | 10:47 |
| Razique | so on boot in many cases, we won't loose time by creating a swapfile | 10:47 |
| *** popux has quit IRC | 10:47 | |
| Razique | livemoon: yah | 10:47 |
| livemoon | provide another swap? | 10:47 |
| Razique | livemoon: nope; the only one | 10:47 |
| Razique | today instances are swapless | 10:47 |
| *** zigo has joined #openstack | 10:47 | |
| *** longhudou has quit IRC | 10:47 | |
| livemoon | oh | 10:47 |
| Razique | so that script creates a swap based on the flavor | 10:48 |
| Razique | and set a 1gb thresholf | 10:48 |
| Razique | threshold | 10:48 |
| livemoon | when I create rhel image,if there is no swap, installation will failed | 10:48 |
| livemoon | Does your script be pushed into your github? | 10:49 |
| tungvs | livemoon: swap partition can be made when installing RHEL, and be tripped off using losetup after that for saing space | 10:49 |
| *** MarkA2 has quit IRC | 10:49 | |
| livemoon | tunavs: tks, I will try | 10:50 |
| tungvs | Razique: maybe I'm the only noob here, but I want to ask how to retrieve flavor through "metadata" in the instance ? | 10:51 |
| tungvs | and what is "metadata" btw ? | 10:51 |
| foexle | an example for metadatas are the used key-pair | 10:55 |
| tungvs | oh I see, that's the "curl http://169.254.169.254/latest/meta-data" step, right | 10:56 |
| tungvs | ok, I will dig deeper, tks foexle | 10:57 |
| *** zigo-_- has joined #openstack | 10:57 | |
| Razique | tungvs: yup it's an Amazon stuff | 10:58 |
| Razique | the api server contains info about the newly created instance | 10:59 |
| Razique | that one connects to it and retrive insfos | 10:59 |
| Razique | public keys mainly | 10:59 |
| tungvs | Razique: is there any document about it ? | 10:59 |
| Razique | so you can ssh to the instance after it booted, without the pain of setting password | 10:59 |
| Razique | sure | 10:59 |
| livemoon | api server? | 10:59 |
| Razique | http://www.google.fr/search?gcx=c&sourceid=chrome&ie=UTF-8&q=amazon+metadatas#sclient=psy-ab&hl=fr&source=hp&q=amazon+metadatas+169.254&pbx=1&oq=amazon+metadatas+169.254&aq=f&aqi=&aql=&gs_sm=e&gs_upl=2950l5261l0l5817l8l7l0l0l0l0l505l1940l0.1.5.5-1l7l0&bav=on.2,or.r_gc.r_pw.r_cp.,cf.osb&fp=64f4fd902b05af1a&biw=1318&bih=955 | 10:59 |
| Razique | http://open.eucalyptus.com/participate/wiki/accessing-instance-metadata | 10:59 |
| *** saju_m has joined #openstack | 11:00 | |
| tungvs | that's great, Razique | 11:00 |
| tungvs | have to go for dinner, good to see you guys. | 11:00 |
| *** zigo has quit IRC | 11:01 | |
| *** zigo-_- is now known as zigo | 11:01 | |
| livemoon | dinner? | 11:01 |
| livemoon | we have the same timezone maybe | 11:02 |
| Razique | you welcome tungvs | 11:02 |
| Razique | later :) | 11:02 |
| *** saju_m has quit IRC | 11:02 | |
| *** saju_m has joined #openstack | 11:03 | |
| *** coli_ has joined #openstack | 11:03 | |
| *** coli has quit IRC | 11:03 | |
| *** saju_m has quit IRC | 11:03 | |
| *** saju_m has joined #openstack | 11:04 | |
| *** saju_m has joined #openstack | 11:05 | |
| Razique | ow man | 11:05 |
| Razique | I did nothing this morning | 11:05 |
| Razique | just been lazy on internet | 11:05 |
| livemoon | :) | 11:06 |
| livemoon | I am busy today | 11:06 |
| livemoon | coding to monitor my VM | 11:07 |
| *** saju_m has quit IRC | 11:07 | |
| Razique | how livemoon ? | 11:07 |
| *** saju_m has joined #openstack | 11:07 | |
| livemoon | using libvirt | 11:07 |
| *** tungvs has quit IRC | 11:07 | |
| *** saju_m has quit IRC | 11:08 | |
| livemoon | I just start to learn | 11:08 |
| *** saju_m has joined #openstack | 11:08 | |
| livemoon | I think your work about metadata is useful for me | 11:08 |
| *** ahasenack has joined #openstack | 11:11 | |
| *** katkee has quit IRC | 11:11 | |
| Razique | sure | 11:16 |
| Razique | i'll see how I could well integrate that extra step | 11:16 |
| livemoon | teach me please | 11:18 |
| Razique | BasTichelaar: here suggested to tweak the api data to be provided | 11:19 |
| Razique | if you can get the ram, you can also get the swap | 11:19 |
| *** GheRivero has quit IRC | 11:23 | |
| *** GheRivero has joined #openstack | 11:25 | |
| *** livemoon has left #openstack | 11:26 | |
| *** titan has joined #openstack | 11:32 | |
| *** popux has joined #openstack | 11:32 | |
| *** miclorb_ has joined #openstack | 11:33 | |
| *** fulcrum has quit IRC | 11:35 | |
| zykes- | notmyname: ? | 11:37 |
| titan | Hi guys, here comes an unsolicited questions: does openstack have a zone wide DNS resolution solution or something similar in the blueprints? At the moment, we rely on dnsmasq to read local file to achieve dns resolution. However, should more than one dnsmasq running in different servers, there seems to be no way to cross servers to do name resolution | 11:40 |
| titan | thx in advance, guys | 11:40 |
| foexle | hey Razique do you use swift ? | 11:44 |
| Razique | foexle: not yet :( | 11:44 |
| foexle | ah ok | 11:44 |
| *** cyberdo has joined #openstack | 11:44 | |
| foexle | but a question .... i found a resolved high critical bug ... this one was resolved in nov. but i cant see a merge in the stable ubuntu packages | 11:45 |
| cyberdo | quick questions: does openstack support automatic migration using local files? Can it use LVM as these local files? | 11:46 |
| *** saju_m has joined #openstack | 11:46 | |
| foexle | https://bugs.launchpad.net/swift/+bug/882679 | 11:46 |
| foexle | normally should the package maintainer resolve this bug in the package version too | 11:46 |
| Razique | cyberdo: what do you mean by "local file" ? | 11:47 |
| cyberdo | Razique: non-shared storage... storage is held on the virtualization nodes | 11:47 |
| *** miclorb_ has quit IRC | 11:48 | |
| Razique | yah both ways | 11:49 |
| Razique | live migration or block storage migration | 11:50 |
| cyberdo | our nodes use very little disk (often about 10G per server), so I rather keep the disk locally... Also it reduces the impact of failures (having a SAN breaking down is not funny) | 11:50 |
| cyberdo | ooh.. nice | 11:50 |
| cyberdo | also: this using xen? | 11:50 |
| Razique | cyberdo: I think it's supported with Xen | 11:50 |
| Razique | here | 11:50 |
| Razique | http://docs.openstack.org/cactus/openstack-compute/admin/content/configuring-live-migrations.html | 11:50 |
| cyberdo | (I saw that some features differed between different hosts) | 11:50 |
| Razique | it's the kvm way | 11:50 |
| cyberdo | thanks a lot | 11:51 |
| Razique | basically nova-manage vm migrate --ec2-id --dest | 11:51 |
| Razique | sorry | 11:51 |
| Razique | nova-manage vm live_migration | 11:51 |
| Razique | and nova-manage vm block_migration | 11:51 |
| titan | Hi guys, here comes an unsolicited questions: does openstack have a zone wide DNS resolution solution or something similar in the blueprints? At the moment, we rely on dnsmasq to read local file to achieve dns resolution. However, should one dnsmasq running in different servers in multi_host mode, there seems to be no way to cross servers to do name resolution | 11:54 |
| cyberdo | Razique: 'Instances: Instance can be migrated with ISCSI/AoE based volumes' ;o/ | 11:55 |
| Razique | cyberdo: what about lvm ? | 11:56 |
| Razique | oops | 11:56 |
| Razique | misread | 11:56 |
| Razique | yah it is said, but never really tried yet | 11:56 |
| Razique | I'll | 11:56 |
| *** nijaba has quit IRC | 11:57 | |
| Razique | titan: zykes- here is the man | 11:58 |
| zykes- | Razique: ? | 11:58 |
| Razique | zykes-: titan has a DNS question :p | 11:59 |
| Razique | Hi guys, here comes an unsolicited questions: does openstack have a zone wide DNS resolution solution or something similar in the blueprints? At the moment, we rely on dnsmasq to read local file to achieve dns resolution. However, should one dnsmasq running in different servers in multi_host mode, there seems to be no way to cross servers to do name resolution | 11:59 |
| titan | zykes: yep, I have one | 11:59 |
| *** dendro-afk is now known as dendrobates | 11:59 | |
| *** zz_bonzay is now known as bonzay | 12:00 | |
| titan | Razique: thx | 12:00 |
| cyberdo | Razique: I suppose I can still look into it... otherwise there might be some distributed iSCSI-server out there | 12:00 |
| zykes- | titan: dns howo ? | 12:00 |
| Razique | cyberdo: I need to validate migration with iscsi | 12:00 |
| Razique | I was afraid the I/O part | 12:01 |
| zykes- | yeah, there was an idea on "dnsaas" | 12:01 |
| zykes- | but I haven't had the time to do anything on it | 12:01 |
| zykes- | atm i'm working on a thing with Swift instead | 12:01 |
| *** nijaba has joined #openstack | 12:01 | |
| Razique | what happens if there are ongoing writes on the instance while you perform the migration | 12:01 |
| *** bonzay is now known as zz_bonzay | 12:01 | |
| titan | zykes: is that okay to say that there is no zone wide DNS service at the moment? | 12:02 |
| zykes- | titan: think so yeah | 12:03 |
| titan | zykes: cool. thx1 | 12:04 |
| zykes- | i'm trying to front swift as a archival service for media files | 12:04 |
| *** nijaba has quit IRC | 12:04 | |
| *** saju_m has quit IRC | 12:05 | |
| BasTichelaar | titan: as far as I know, openstack does nothing with dns | 12:09 |
| BasTichelaar | hmm, didnt scroll to the bottom :) | 12:09 |
| zykes- | http://etherpad.openstack.org/HkEvt4crw9 | 12:10 |
| zykes- | check that out | 12:10 |
| *** nijaba has joined #openstack | 12:17 | |
| *** nijaba has joined #openstack | 12:17 | |
| *** supriya has quit IRC | 12:19 | |
| *** ben_duyujie has joined #openstack | 12:21 | |
| *** shawn_ is now known as shawn | 12:29 | |
| *** mcclurmc has quit IRC | 12:31 | |
| *** mcclurmc has joined #openstack | 12:33 | |
| *** ben_duyujie has quit IRC | 12:35 | |
| coli_ | razique: what do you mean by writes ? | 12:36 |
| *** bsza has joined #openstack | 12:37 | |
| *** ben_duyujie has joined #openstack | 12:37 | |
| uvirtbot | New bug: #899116 in keystone "The wrong keystone module is used when building docs" [Undecided,In progress] https://launchpad.net/bugs/899116 | 12:37 |
| *** po has joined #openstack | 12:42 | |
| *** n8 has joined #openstack | 12:42 | |
| *** n8 is now known as Guest3171 | 12:43 | |
| *** pradeep1 has quit IRC | 12:52 | |
| *** ben_duyujie has left #openstack | 12:52 | |
| *** markvoelker has joined #openstack | 12:56 | |
| *** livemoon has joined #openstack | 12:58 | |
| *** PeteDaGuru has joined #openstack | 13:04 | |
| *** mcclurmc has quit IRC | 13:04 | |
| *** mcclurmc has joined #openstack | 13:05 | |
| zykes- | notmyname: here ? | 13:08 |
| *** yshh has joined #openstack | 13:11 | |
| *** PeteDaGuru has quit IRC | 13:12 | |
| *** yshh has joined #openstack | 13:16 | |
| *** lts has joined #openstack | 13:17 | |
| *** dolphm has joined #openstack | 13:17 | |
| *** egant has quit IRC | 13:18 | |
| notmyname | zykes-: morning | 13:20 |
| *** osier has quit IRC | 13:21 | |
| *** javiF has quit IRC | 13:21 | |
| Spirilis | ok question- When using "nova-manage network create ...", it looks like it completes without error (and /var/log/nova/nova-manage.log looks good), but no entries are added to the DB ... what's the first thing I should check? | 13:23 |
| Spirilis | The /etc/nova/nova.conf file has the correct --sql_connection string | 13:23 |
| *** lloydde has joined #openstack | 13:23 | |
| *** abecc has joined #openstack | 13:24 | |
| Spirilis | and it looks like the DB user specified in there does have GRANT ALL PRIVILEGES access to that db | 13:24 |
| Spirilis | fyi this is using Ubuntu's natty repo copy of openstack, minus the keystone piece which I git'd | 13:25 |
| Spirilis | er | 13:27 |
| Spirilis | Oneiric repo, not natty | 13:27 |
| Spirilis | using ubuntu 11.10 here | 13:27 |
| Spirilis | can't keep track of their crazy names ;) | 13:27 |
| foexle | anyone knows how i can change password of an user with keystone-manage ? | 13:27 |
| zykes- | gawd this company and opensource | 13:30 |
| zykes- | funny thing | 13:30 |
| Razique | back | 13:32 |
| *** Razique has left #openstack | 13:32 | |
| *** mcclurmc has quit IRC | 13:33 | |
| *** mcclurmc has joined #openstack | 13:34 | |
| *** mjfork has joined #openstack | 13:34 | |
| *** lorin1 has joined #openstack | 13:34 | |
| *** Razique has joined #openstack | 13:35 | |
| zykes- | notmyname: can you go from having 2 nodes with swift with 2 zones on each | 13:36 |
| zykes- | to having 4 nodes + with 1 zone pr box or rack ? | 13:36 |
| Spirilis | zykes-: yes, if you are trying to move from a "2 zones/node" to "1 zone/node" config it might be prudent to add the new nodes first, rebalance the rings once or twice, let it sync up, then remove one of the extra zones from the original 2 servers, rebalance/let it resync, then re-add that disk under the same zone as the one already on the box | 13:38 |
| *** Tristani3D has joined #openstack | 13:38 | |
| Tristani3D | Hello | 13:38 |
| Tristani3D | I got a question | 13:39 |
| Tristani3D | I am following the Starter Guide but I am running into an issue | 13:39 |
| Tristani3D | on the service iscsitarget restart part | 13:39 |
| Tristani3D | I receive the error: FATAL: Module iscsi_targt not found | 13:40 |
| Tristani3D | any ideas? | 13:40 |
| Tristani3D | ps. I have installed isxcsitarget-dkms | 13:40 |
| mjfork | Tristani3D: on Ubunutu? | 13:40 |
| Tristani3D | package | 13:40 |
| Tristani3D | yes | 13:40 |
| mjfork | hmm - guessing here - do you have nova-voume installed? | 13:41 |
| mjfork | can you pastebin the full trace | 13:41 |
| Tristani3D | nova-volume is indead part of the nova install section | 13:41 |
| Tristani3D | ill pastebin the output, 1sec. | 13:43 |
| notmyname | zykes-: what Spirilis said | 13:44 |
| Tristani3D | @mjfork http://pastebin.com/DqPbdNTC | 13:47 |
| larissa | Tristani3D: Error: "mjfork" is not a valid command. | 13:47 |
| zykes- | ok | 13:47 |
| zykes- | but i guess i'll never the the light of swift here anyways, people here want a more conventional storage option for a media archive :p | 13:48 |
| notmyname | zykes-: sounds frustrating | 13:48 |
| Tristani3D | mjfork: http://pastebin.com/DqPbdNTC | 13:48 |
| zykes- | notmyname: yeah, kindof | 13:48 |
| mjfork | Tristani3D: so it is NOT openstack throwing this error? | 13:48 |
| notmyname | zykes-: any arguments I can add that would help? | 13:49 |
| zykes- | notmyname: i dunno, a person here wants it to be cheap so :p | 13:49 |
| *** mnabil has joined #openstack | 13:49 | |
| zykes- | and i guess 30k $ isn't ckeap enough for a 24tb deployment | 13:49 |
| Tristani3D | mjfork: I have no clue | 13:49 |
| mjfork | so, modprobe is failing oot ;ppls ;ole | 13:50 |
| mjfork | woops, modprobe is failing it looks like | 13:50 |
| notmyname | zykes-: 1) scales more than conventional storage 2) since the replicas/self-healing nature you don't have to have separate hardware for backups 3) can use commodity hardware (saving money) | 13:50 |
| mjfork | Tristani3D: http://sysadmin.wikia.com/wiki/Server_side_iscsi_linux | 13:50 |
| zykes- | notmyname: but what about off site ? | 13:51 |
| zykes- | like in case of a flood or so :p | 13:51 |
| Tristani3D | mjfork: well, strange, I have followed every line of the guide ;) | 13:51 |
| mjfork | yeah, unknown to me. what version of ubunutu? | 13:51 |
| zykes- | notmyname: the solution we're going for now is 2*24tb dell boxes with windows + veeam backup | 13:52 |
| Tristani3D | 11.10 | 13:52 |
| zykes- | the 2nd box will replicate the first box over a wan link | 13:52 |
| notmyname | zykes-: ya, you could either have a 2nd cluster and sync or you could do something like 4 zones (2 in each location). with 3 replicas you are guaranteed at least one copy in both locations | 13:52 |
| mjfork | Tristani3D: i don't know, but that should fix you up | 13:52 |
| *** dprince has joined #openstack | 13:52 | |
| Tristani3D | mjfork: Ill give it a go, and let you know. Thx | 13:52 |
| *** zz_bonzay is now known as bonzay | 13:52 | |
| zykes- | notmyname: but say for trunked objects then, will 1 replica be a full copy of all chunks ? | 13:52 |
| *** bonzay is now known as zz_bonzay | 13:53 | |
| Tristani3D | mjfork: meh, http://debian.hug.cx/debian/ is 404 :( | 13:54 |
| *** mcclurmc has quit IRC | 13:54 | |
| *** GheRivero has quit IRC | 13:54 | |
| notmyname | zykes-: no, each chunk will be in 2 of the 4 locations. if you always have 4 zones, you will always have at least 1 copy of each entity in both places. you'd have to be aggressive about hardware failures though (eg HDD with the only copy in location A goes down and may not be restored before location B goes down) | 13:54 |
| notmyname | zykes-: s/2 of the 4/3 of the 4/ | 13:55 |
| *** mcclurmc has joined #openstack | 13:55 | |
| zykes- | notmyname: doesn't each chunk have 3 replicas ? :| | 13:55 |
| notmyname | ya | 13:55 |
| *** yshh has quit IRC | 13:55 | |
| zykes- | what you mean by being aggressive then ? | 13:55 |
| *** objectiveous has quit IRC | 13:56 | |
| *** katkee has joined #openstack | 13:57 | |
| mjfork | Tristani3D: what is that? i think you just need to run those 3 commands | 13:59 |
| mjfork | Tristani3D: apt-get then m-a then modprobe iscsitarget | 13:59 |
| mjfork | just skip down to installing the packages | 14:00 |
| *** mnabil has quit IRC | 14:01 | |
| Tristani3D | mjfork: Done, seems to work now | 14:04 |
| Razique | argh (nova.compute.manager): TRACE: Stderr: 'ERROR: trying to add VLAN #100 to IF -:None:- error: No such device\n' | 14:04 |
| *** katkee has quit IRC | 14:04 | |
| Tristani3D | mjfork: iscsi-source seems to be the trick and missing package | 14:04 |
| Razique | bridge_interface=eth0 in nvoa.conf and NULL into db ; any hint ? | 14:04 |
| Tristani3D | mjfork: might be a good one to add this in the docs | 14:05 |
| *** hadrian has joined #openstack | 14:05 | |
| *** AlanClark has joined #openstack | 14:05 | |
| Razique | mjfork: can you confirm on my comment ? | 14:06 |
| Razique | if it's working, i'll update the doc so | 14:06 |
| zykes- | can't the damned prices go down for disks soon | 14:07 |
| Razique | http://rww.readwriteweb.netdna-cdn.com/cloud/SHI%20blog%20Cloudville%20-%20Glue.jpg | 14:07 |
| Razique | zykes-: yh :/ | 14:07 |
| mjfork | Razique: what commments? | 14:07 |
| mjfork | Tristani3D: what guide where you following? | 14:07 |
| zykes- | i might as well give up this fight, useless :p | 14:07 |
| Tristani3D | mjfork: https://bugs.launchpad.net/openstack-manuals/+bug/868595?comments=all | 14:07 |
| Razique | mjfork: weren't you talking about nova-volume setup issue | 14:07 |
| Razique | regarding the iscsi part | 14:08 |
| Tristani3D | mjfork: "Openstack Compute Starter Guide" Nov 11, 2011 | 14:08 |
| livemoon | hi, Razique | 14:08 |
| mjfork | looks like Razique documented the bug that Tristani3D ran into at that link | 14:09 |
| livemoon | how to use your swap scripte when booting an instance? | 14:09 |
| mjfork | and the fix | 14:09 |
| foexle | livemoon: you need to add this script in the startup area | 14:10 |
| livemoon | foexle: how to use in novaclient command? | 14:10 |
| foexle | novaclient? .... not with novaclient | 14:11 |
| notmyname | zykes-: just that in my example there is a slight risk of data loss if the one copy in one location is lost at the same time the entire other location is lost | 14:11 |
| foexle | you need change you image startup | 14:11 |
| Razique | Tristani3D: ok great | 14:11 |
| Razique | I'll update the doc | 14:11 |
| foexle | +to | 14:11 |
| Razique | i'm looking into the source file and reduce the line size | 14:11 |
| livemoon | foexle: teach me | 14:11 |
| *** rsampaio has joined #openstack | 14:12 | |
| zykes- | if the disk prices had been what they where earlier | 14:12 |
| livemoon | add into rc.local? | 14:12 |
| zykes- | one simple swift deployment would cost only 15520$ | 14:12 |
| Razique | yah :/ | 14:12 |
| Razique | zykes-: ramdisk ? | 14:12 |
| Razique | :p | 14:12 |
| zykes- | Razique: no, 2tb sata prices | 14:12 |
| Glacee | lol ramdisk is still expensive :) | 14:13 |
| foexle | livemoon: which distro ? | 14:13 |
| Glacee | even for vm load | 14:13 |
| zykes- | if you work out of the price increase % and subtract that from todays disk prices | 14:13 |
| livemoon | rhel | 14:13 |
| zykes- | and take todays server prices without disks | 14:13 |
| notmyname | zykes-: how big is that deployment? | 14:13 |
| livemoon | Razique: I think nova-volume docs is old | 14:13 |
| livemoon | if we use 11.10, do not use iscsitarget | 14:14 |
| zykes- | notmyname: 4 boxes of 24 tb notmyname | 14:14 |
| Tristani3D | Razique: btw, the lines are not in the starter guide its seems | 14:14 |
| Tristani3D | Razique: but in the OpenStack Compute Administration Manual | 14:15 |
| Tristani3D | Razique: page 103 | 14:15 |
| Glacee | have anyone tried the new Seagate 3Tb drive that came out last week? Seems pretty neat | 14:15 |
| Tristani3D | Razique: and the image on page 85 is bugged as well, way to big for an A4 sized page | 14:16 |
| coli_ | tristani3D: docs are for developers, however your best documentation is in the source code, | 14:16 |
| Glacee | zykesL: with which brand did you finally go with for your seervers? | 14:16 |
| *** rsampaio has quit IRC | 14:16 | |
| coli_ | tristani3D: then correct that bugged page :-) it's open source | 14:16 |
| Razique | Tristani3D: don't hesistate to open doc bugs for things you notice | 14:17 |
| Razique | I regularly track doc bugs and try to fix. dispatch em | 14:17 |
| zykes- | Glacee: Dell R515 | 14:17 |
| zykes- | came out the cheapest actually | 14:17 |
| Tristani3D | coli: Razique: Thanks, im new to this. No idea how and where to edit :) | 14:18 |
| Glacee | ok cool | 14:18 |
| Razique | Tristani3D: https://bugs.launchpad.net/openstack-manuals/ | 14:18 |
| Glacee | cheapeast than SM? | 14:18 |
| zykes- | but i'm getting sold out to a customer anyways, tired of not getting ideas through here so might as well help someone else | 14:18 |
| Razique | click "Report a bug" | 14:18 |
| Razique | and we'll see it | 14:18 |
| Glacee | hehe ok | 14:18 |
| *** bryguy has quit IRC | 14:18 | |
| *** danishman has joined #openstack | 14:18 | |
| Tristani3D | Razique: ok | 14:18 |
| *** mrjazzcat has joined #openstack | 14:19 | |
| *** nelson1234 has quit IRC | 14:21 | |
| *** bryguy has joined #openstack | 14:21 | |
| *** doude has joined #openstack | 14:21 | |
| doude | Hi stackers | 14:22 |
| Razique | oi doude :) | 14:23 |
| doude | Who use the Diablo release with libvirt KVM on Ubuntu 11.10 ? | 14:23 |
| Tristani3D | coli_: Razique: filled those two bugs, thx | 14:23 |
| *** sgg has joined #openstack | 14:23 | |
| doude | Razique: hello | 14:23 |
| Razique | Tristani3D: perfect | 14:23 |
| *** lvaughn has joined #openstack | 14:24 | |
| *** dolphm has quit IRC | 14:24 | |
| doude | I get libvirt freeze evry 24 hours. Do you met that problem ? | 14:25 |
| zykes- | notmyname: how come ? | 14:25 |
| notmyname | zykes-: quick calculations from newegg and dell websites: r815 (2800) + 2 supermicro JBODs (4400) + 90 2TB HDDs (90*255) / 90000 (marketing GB) / 4 (marketing to billable conversion) = hardware corst of $.08 per billable GB | 14:25 |
| *** mrjazzcat has left #openstack | 14:26 | |
| notmyname | zykes-: how come what? | 14:26 |
| Razique | doude: any error , | 14:26 |
| *** gnu111 has joined #openstack | 14:27 | |
| doude | Razique: strange, I've got it on 3 differents hosts | 14:27 |
| Razique | anyone have met ERROR: trying to add VLAN #100 to IF -:None:- error: No such device\n ? | 14:27 |
| doude | Do you use Ubuntu 11.10 ? | 14:27 |
| Razique | it's fricking weird :/ | 14:27 |
| Razique | 10.04 | 14:27 |
| Razique | Lucid | 14:28 |
| doude | and libvirt version ? | 14:28 |
| *** lborda has joined #openstack | 14:28 | |
| Razique | 0.8.8-1ubuntu3~ppalucid1 | 14:28 |
| *** rsampaio has joined #openstack | 14:28 | |
| Razique | in fact i'm amazed, since It was working fine | 14:28 |
| doude | mine libvirt 0.9.2 | 14:29 |
| Razique | the subtility here is that my node uses eth0 while the controller eth1 | 14:29 |
| Razique | doude: can you check syslog ? | 14:29 |
| zykes- | notmyname: why the 90*255 ? | 14:29 |
| doude | Nothing in syslog | 14:30 |
| notmyname | zykes-: HDD cost on newegg | 14:30 |
| notmyname | to fill up the jbods | 14:30 |
| zykes- | yeah, but then comes the hw cost in addition or ? | 14:30 |
| Razique | doude: nova-compute.log ? | 14:30 |
| doude | I just found this in libvirt logs : http://paste.openstack.org/show/3605/ | 14:32 |
| zykes- | notmyname: what disks is that ? | 14:32 |
| doude | Razique: nova-compute wait a response from libvirt and the compute service is frozen | 14:32 |
| foexle | doude: is you database inconsistent ? | 14:33 |
| doude | if I restart the libvirtd daemon, the compute restart itself | 14:33 |
| doude | foexle: no the db stay clear | 14:33 |
| doude | evrything work fine after I restarted the libvirt daemon | 14:34 |
| notmyname | zykes-: http://www.newegg.com/Product/Product.aspx?Item=N82E16822148506 | 14:34 |
| zykes- | why you need 90 of them ? | 14:34 |
| doude | some use libvirt 0.9.2 ? | 14:34 |
| *** dolphm has joined #openstack | 14:34 | |
| foexle | how many instances are running ? | 14:34 |
| notmyname | zykes-: 2 45-slot JBDS | 14:34 |
| zykes- | that's not the dell r815 ? | 14:35 |
| foexle | doude: i'm using 0.9.2 | 14:35 |
| doude | foexle: and no problem ? | 14:35 |
| foexle | doude: nope .... no problems like this ;) | 14:36 |
| notmyname | zykes-: http://www.newegg.com/Product/Product.aspx?Item=N82E16811152143 http://www.dell.com/us/business/p/poweredge-r815/pd?refid=poweredge-r815&st=dell%20PowerEdge%20r815&dgc=ST&cid=76607&lid=1971537&acd=sEDGQGCI0,17346158020,901qz26673&isredir=true | 14:36 |
| notmyname | zykes-: dell.com says $2800 | 14:36 |
| foexle | do you have checkd your hdd's with smart or controller bios ? | 14:36 |
| Tristani3D | I have a general question, am I right that Openstack doens't rely on SAN storage but distribute its storage to all disks in every single node instead? | 14:36 |
| zykes- | what's the r815 for ? | 14:37 |
| notmyname | zykes-: the controller for the 2 supermicro JBODS | 14:37 |
| ashp | Tristani3D: I think, and I'm just an idiot that wants to use openstack eventually myself, you can do it either way. SAN or distributed filesystem of some kind. | 14:37 |
| notmyname | gotta run. be back in about an hour (ish) | 14:38 |
| Glacee | notmyname: do you think its less expensive.. buy a small 1u server + jobd.. or buying the monster 4U server with 36 slots | 14:38 |
| Tristani3D | ashp: Right, without the use of a SAN its going to be much cheaper.. | 14:38 |
| doude | foexle: Ok thanks. I'll try to found a way to reproduce that | 14:38 |
| *** yshh has joined #openstack | 14:39 | |
| *** gavri has joined #openstack | 14:39 | |
| foexle | doude: yeah sorry :) | 14:39 |
| ashp | Tristani3D: That's one reason I'm interested - obtaining shared storage here is a nightmare | 14:41 |
| zykes- | notmyname: we don't want jbods | 14:42 |
| *** ryker has joined #openstack | 14:42 | |
| zykes- | sorry notmyname i meant SM | 14:42 |
| *** oubiwann has joined #openstack | 14:48 | |
| *** darthflatus has joined #openstack | 14:49 | |
| *** eugene has joined #openstack | 14:51 | |
| Tristani3D | ashp: yeah, i like to concept of using all the disk in all nodes instead of one (or multiple) storage instance | 14:51 |
| *** eugene is now known as Guest36102 | 14:51 | |
| *** Guest36102 is now known as eugenes | 14:52 | |
| *** rustam has quit IRC | 14:54 | |
| *** ryker has quit IRC | 14:55 | |
| *** ryker has joined #openstack | 14:55 | |
| zykes- | notmyname: does RS today do cross site replicating ? | 14:56 |
| *** aliguori has joined #openstack | 14:58 | |
| *** rustam_ has joined #openstack | 14:59 | |
| *** dgags has joined #openstack | 15:05 | |
| *** mjfork has quit IRC | 15:06 | |
| *** katkee has joined #openstack | 15:06 | |
| uvirtbot | New bug: #899186 in openstack-manuals "Line size bugged page 103" [Undecided,New] https://launchpad.net/bugs/899186 | 15:07 |
| Razique | regarding nova-flags | 15:08 |
| Razique | they either reside in database or nova.conf ? | 15:08 |
| uvirtbot | New bug: #899189 in openstack-manuals "Image to large at page 85 of OpenStack Compute Administration Manual" [Undecided,New] https://launchpad.net/bugs/899189 | 15:08 |
| *** katkee has quit IRC | 15:11 | |
| btorch | zykes-: he is probably on the way to the office, but we don't do cross DC replication if that's what you mean | 15:11 |
| *** mjfork has joined #openstack | 15:11 | |
| zykes- | btorch: so one can't have data across datacenters ? | 15:11 |
| btorch | zykes-: Not saying that you can't, swift has container sync and there is nothing really stopping one from having a zone in different location. You need to make sure you have a good pipe to reach that location | 15:12 |
| btorch | zykes-: like a dedicated link for example | 15:13 |
| *** imsplitbit has joined #openstack | 15:14 | |
| *** code_franco has joined #openstack | 15:15 | |
| *** ryker_ has joined #openstack | 15:16 | |
| Tristani3D | Ok I am having this next issue | 15:18 |
| Tristani3D | root@ubuntu:/home/localadmin# nova-manage db sync | 15:18 |
| Tristani3D | Command failed, please check log for more info | 15:18 |
| Tristani3D | Any idea, which logs? | 15:18 |
| *** clauden_ has quit IRC | 15:18 | |
| *** clauden has joined #openstack | 15:18 | |
| *** ryker has quit IRC | 15:19 | |
| *** ryker_ is now known as ryker | 15:19 | |
| *** Rajaram has quit IRC | 15:20 | |
| *** clauden__ has joined #openstack | 15:21 | |
| *** clauden has quit IRC | 15:21 | |
| *** clauden__ is now known as clauden | 15:21 | |
| *** eugenes has left #openstack | 15:21 | |
| gnu111 | Tristani3D: check nova-api.log, I think. or nova-manage.log. | 15:22 |
| Glacee | zykes: You can have multiple datacenter in the same cluster/ring | 15:23 |
| Glacee | You need to makre sure that you have a low latency line between both sites and that you seperate your failure zones properly | 15:23 |
| Glacee | it will slow down swift.. if your link is the bottleneck | 15:24 |
| Glacee | dont do it over the internet for an example :0 | 15:25 |
| *** dubsquared has joined #openstack | 15:25 | |
| *** deshantm_laptop has joined #openstack | 15:25 | |
| Tristani3D | gnu111: thx, I see (nova): TRACE: OperationalError: (OperationalError) (1045, "Access denied for user 'novadbadmin'@'192.168.56.130' (using password: YES)") None None | 15:25 |
| *** cmagina has quit IRC | 15:25 | |
| *** GheRivero has joined #openstack | 15:25 | |
| *** kbringard has joined #openstack | 15:25 | |
| Tristani3D | gnu111: might have been a misconfiguration somewhere | 15:25 |
| *** cmagina has joined #openstack | 15:25 | |
| gnu111 | Tristani3D: is the mysql db started? also check the firewall. make sure the compute node can talk to the db. | 15:26 |
| Tristani3D | gnu111: mysql 2331 0.0 2.4 174400 24980 ? Ssl 06:52 0:00 /usr/sbin/mysqld | 15:26 |
| Tristani3D | gnu111: mysql is running | 15:27 |
| dubsquared | can you log into the db with the username and password specified in the nova.conf? | 15:27 |
| *** qazwsx has joined #openstack | 15:27 | |
| gnu111 | is 192.168.56.130 the compute node or the controller? | 15:28 |
| *** primeministerp has joined #openstack | 15:28 | |
| *** CaptTofu has quit IRC | 15:28 | |
| *** sgg has quit IRC | 15:28 | |
| Tristani3D | gnu111: controller, im following the compute starter guide section "server 1" | 15:28 |
| *** ldlework has joined #openstack | 15:29 | |
| Tristani3D | dubsquared: ERROR 1045 (28000): Access denied for user 'novadbadmin'@'localhost' (using password: YES) | 15:30 |
| Tristani3D | root@ubuntu:/home/localadmin# | 15:30 |
| notmyname | zykes-: sorry. my math was a little wrong earlier. I should have divided by 180000 (for total marketing GB rather than 90000) | 15:30 |
| notmyname | zykes-: (2800 + 4400 + (90*255)) / 180000. / 4 = $0.041875 / GB | 15:30 |
| Sander^work | Anyone around that have got horizon with keystone working? | 15:30 |
| gnu111 | Tristani3D: there you go. did you do something like this for the db: GRANT ALL PRIVILEGES ON $DB_NAME.* TO $DB_USER IDENTIFIED BY '$DB_PASS ? | 15:31 |
| Tristani3D | gnu111: mysql -uroot -pmygreatsecret -e "SET PASSWORD FOR 'novadbadmin'@'%' = | 15:31 |
| Tristani3D | PASSWORD('novasecret');" | 15:31 |
| Razique | soren: here ? | 15:31 |
| soren | Razique: Kinda. Gimme 5 minutes. | 15:31 |
| Tristani3D | gnu111: direct copy of the docs, but due to the enter it fails to execute proparly | 15:31 |
| Razique | soren: sure thanks | 15:31 |
| Tristani3D | gnu111: root@ubuntu:/home/localadmin# nova-manage db sync | 15:32 |
| Tristani3D | MigrateDeprecationWarning | 15:32 |
| Tristani3D | works now | 15:32 |
| *** katkee has joined #openstack | 15:32 | |
| Tristani3D | or not? | 15:32 |
| *** reed has joined #openstack | 15:34 | |
| *** ldleworker has joined #openstack | 15:34 | |
| *** ldlework has quit IRC | 15:35 | |
| notmyname | zykes-: oops. it was pointed out to me that I should have multiplied by 4 rather than divide by 4 | 15:35 |
| notmyname | zykes-: that makes it $.67 per GB | 15:35 |
| notmyname | much closer (but still lower) than your specs | 15:35 |
| *** katkee has quit IRC | 15:37 | |
| soren | Razique: 'sup? | 15:38 |
| Razique | fine and you , | 15:39 |
| Razique | . | 15:39 |
| Razique | ? erf :D | 15:39 |
| Razique | ok I've a a issue here with that file | 15:40 |
| Razique | http://pastie.org/2955419 | 15:40 |
| *** livemoon has left #openstack | 15:40 | |
| Razique | usr/lib/pymodules/python2.6/nova/network/linux_net.py | 15:40 |
| gnu111 | notmyname: do you know if cyberduck works with the tempauth setup in swift? I have only been using the command line so far. | 15:40 |
| Razique | it get the bridge_interface from the database | 15:40 |
| notmyname | gnu111: it should (but I haven't done it myself) | 15:41 |
| Razique | I've tried to add a flag 'bridge_interface'" so it uses nova.conf | 15:41 |
| coli_ | razique: what networking mode are you working in ? | 15:41 |
| Razique | my issue here is that i've a node which doesn't use the same interface as the others | 15:41 |
| Razique | coli_: vlan | 15:41 |
| Razique | so a flag would override the default eth0 | 15:41 |
| coli_ | razique: why woyld you use bridge_interface then for vlan mode ? | 15:41 |
| *** dolphm has quit IRC | 15:41 | |
| *** jaypipes has quit IRC | 15:42 | |
| Razique | coli_: you don't need it ? | 15:42 |
| Razique | My vlan doesn't up | 15:42 |
| coli_ | razique: will check, just a sec | 15:42 |
| Razique | Stderr: 'ERROR: trying to add VLAN #100 to IF -:None:- error: No such device\n' | 15:42 |
| *** primeministerp has quit IRC | 15:42 | |
| Razique | None because the database doesn't contain a bridge_interface | 15:42 |
| Razique | I'm using diablo | 15:43 |
| coli_ | razique: just checked to be sure | 15:43 |
| Razique | the bridge is up on the vlan interface | 15:43 |
| coli_ | you need to use vlan_interface, on that interface it's going to create vlan interfaces, remember that vlan mode create one vlan for each tenant | 15:43 |
| Razique | yup I've the flag vlan_interface :) | 15:44 |
| *** wilmoore has joined #openstack | 15:44 | |
| Razique | but I looked into the code (the file I pasted) | 15:44 |
| Razique | the vlan uses the bridge_interface | 15:44 |
| Razique | _execute('vconfig', 'add', bridge_interface, | 15:44 |
| Razique | and bridge_interface is get from the database | 15:44 |
| coli_ | if you set vlan_interface to eth3 and tenant is got vlan 155 assigned then nova-network will create vlan155 as vlan interface of eth1 and bridge br155 with vlan155 and interfaces from instances | 15:45 |
| coli_ | razique: for flat and flatdhcp | 15:45 |
| Razique | yah | 15:45 |
| Razique | what about vlan mode the ? | 15:45 |
| Razique | why does it use a bridge ? | 15:46 |
| coli_ | just looking into the code, what line was it ? | 15:46 |
| Razique | 892 on my paste | 15:46 |
| *** dolphm has joined #openstack | 15:46 | |
| Razique | http://pastie.org/2955419 | 15:46 |
| Razique | look around line 860 also | 15:46 |
| Razique | network['bridge_interface'], | 15:46 |
| Razique | it would be ideal if we defined a new flag | 15:47 |
| Razique | eg flags.DEFINE_string('bridge_interface','eth0', 'blah') | 15:47 |
| Razique | but I don't know how to "enable it " | 15:47 |
| *** Tristani3D has quit IRC | 15:48 | |
| kbringard | the flags object gets read in, and handles reading it from the config file | 15:48 |
| *** jj0hns0n has joined #openstack | 15:48 | |
| kbringard | so if you added that flag, you would enable it by just setting ābridge_interface in the nova.conf | 15:48 |
| Razique | oi kbringard :) | 15:48 |
| Razique | kbringard: I tried, no luck | 15:48 |
| kbringard | hmmm, I'd have to look at itā¦ I have not added a new flag since before the cactus release, hehe | 15:49 |
| *** vicelow has joined #openstack | 15:49 | |
| kbringard | so maybe the mechanism changed | 15:49 |
| Razique | I also replaced " network['bridge_interface'], | 15:49 |
| Razique | " | 15:49 |
| Razique | by "FLAGS.bridge_interface" | 15:49 |
| vicelow | hello all | 15:49 |
| Razique | but no luck | 15:49 |
| Razique | hey vicelow | 15:50 |
| coli_ | flags.DEFINE_string('vlan_interface', None, 'vlans will bridge into this interface if set') | 15:50 |
| coli_ | from manager.py | 15:50 |
| Razique | coli_: oh, so i's not the "bridge_interface" fielsd ? | 15:50 |
| Razique | field into database* | 15:51 |
| Razique | because It's set into my nova.conf --vlan_interface=eth2 | 15:51 |
| uvirtbot | New bug: #899225 in nova "EC2API: rescue instance can't work" [Undecided,In progress] https://launchpad.net/bugs/899225 | 15:51 |
| coli_ | the vlan's should be created on eth2 and bridge into brNNN where NNN is the vlan id | 15:51 |
| Razique | coli_: yah | 15:52 |
| Razique | weird then my flag is not read :/ | 15:52 |
| Razique | while it was under Cactus | 15:52 |
| Razique | I didn't noticed it earlier cuz cactus had set the vlan properly | 15:53 |
| Razique | and if it exists, it's not recreated | 15:53 |
| coli_ | did you restart nova-network ? | 15:53 |
| Razique | but here I rebooted the lab and saw that | 15:53 |
| Razique | yah | 15:53 |
| Razique | and nova-compute | 15:54 |
| *** rnorwood has joined #openstack | 15:54 | |
| coli_ | check your networks table if it's got correct vlan's bridges set | 15:54 |
| *** popux has quit IRC | 15:54 | |
| Razique | coli_: which field into networks table ? | 15:54 |
| *** dragondm has joined #openstack | 15:54 | |
| *** dragondm has quit IRC | 15:55 | |
| *** dragondm has joined #openstack | 15:55 | |
| coli_ | razique: all of the table if it looks right | 15:55 |
| coli_ | razique: manager.py has all the paremeters defined and explained at the begining of the file | 15:55 |
| Razique | http://paste.openstack.org/show/3609/ | 15:56 |
| *** wilmoore has quit IRC | 15:57 | |
| Razique | coli_: the manager deals with the network table doesn't it ? | 15:57 |
| Razique | it looks like you can't have different interfaces for vlans | 15:57 |
| Razique | eg one server eth3 another one eth2 | 15:57 |
| coli_ | looks fine to me (your paste) | 15:57 |
| Razique | since I don't see where the flag bridge_interface doesn't exist | 15:57 |
| coli_ | what do you mean you cannot have different interfaces for vlans ? | 15:57 |
| Razique | coli_: nova-network on server 1 uses eth0 | 15:58 |
| Razique | nova-compute on server 2 uses eth3 | 15:58 |
| coli_ | on each nova-network host you can have different bridge | 15:58 |
| Razique | I don't use multi_host | 15:58 |
| coli_ | do you have nova-network on each compute host ? | 15:58 |
| Razique | nope one nova-network atm | 15:58 |
| coli_ | so just one nova-network providing info for all compute hosts ? | 15:58 |
| Razique | yah | 15:59 |
| Razique | and on every node a nova-compute only | 15:59 |
| coli_ | in that situation you are right and you can only have one interface :-) | 15:59 |
| Razique | ok so If I install nova-network on the compute node | 15:59 |
| Razique | the flag "bridge_interface" and vlan_interface will work ? | 15:59 |
| Razique | coli_: oh I think I get it | 16:00 |
| *** robbiew has joined #openstack | 16:00 | |
| *** pradeep has joined #openstack | 16:00 | |
| coli_ | you need just vlan_interface if you have nova-network on each host | 16:00 |
| coli_ | I would also strongly suggest nova-api on each host | 16:00 |
| Razique | coli_: why is that ? | 16:00 |
| Razique | coli_: ok leme try :D | 16:01 |
| *** adjohn has joined #openstack | 16:01 | |
| coli_ | razique: look how packets travel to 169.254.169.254 when starting instance and you will see :-) | 16:02 |
| *** uksysadmin has quit IRC | 16:02 | |
| *** kaigan_ has quit IRC | 16:02 | |
| coli_ | razique: if you are going to have separate interface just for that traffic then fine you can have just one nova-api, however if you use public interface then you have a problem :-) | 16:02 |
| *** javiF has joined #openstack | 16:03 | |
| *** imsplitbit has quit IRC | 16:03 | |
| *** javiF has joined #openstack | 16:03 | |
| *** mindpixel has quit IRC | 16:03 | |
| *** cp16net has joined #openstack | 16:04 | |
| Razique | coli_: the api is not publically exposed | 16:05 |
| Razique | coli_: you are the man here | 16:05 |
| Razique | it's working | 16:05 |
| Razique | the reason why I didn't set up multi_host before | 16:05 |
| Razique | is bcause I found out that it was avoiding a live migration Cactus -> Diablo | 16:06 |
| Razique | live : without service interruption | 16:06 |
| *** cctrieloff has joined #openstack | 16:06 | |
| *** reidrac has left #openstack | 16:06 | |
| coli_ | razique: it;s not to do with publicly exposed api but with snat and dnat which is done on the compute host. | 16:06 |
| Razique | ah ? | 16:06 |
| coli_ | currenlty on remote host, how is starting instance communication with api (169.254.169.254) ? | 16:07 |
| Razique | actually the compute node uses the prerouting route in order to access nova-api | 16:08 |
| *** helfrez has joined #openstack | 16:09 | |
| Razique | coli_: in fact maybe I did it wrong the last time | 16:09 |
| Razique | with a running cloud, can I enable/ disable multi_hosting ? | 16:10 |
| coli_ | razique: it's using preroute and then that packet is routed via public interface to nova-api or thorugh other ? | 16:10 |
| Razique | does multi_hosting is mandatory when I've one nova-network per nova-compute also | 16:10 |
| Razique | let me check | 16:10 |
| *** freeflyi1g has joined #openstack | 16:10 | |
| coli_ | razique: I would think so (no idea, just guessing) | 16:10 |
| Razique | coli_: 101 6060 DNAT tcp -- * * 0.0.0.0/0 169.254.169.254 tcp dpt:80 to:172.16.40.11:8773 | 16:11 |
| Razique | ok so it's a local prerouting | 16:12 |
| *** bencherian has joined #openstack | 16:12 | |
| Razique | that hist the nova-api directly | 16:12 |
| Razique | hits* | 16:12 |
| Razique | coli_: that's the thing I don't get in fact regarding diablo | 16:13 |
| coli_ | and then it uses public interface to get to 172.16.40.11 ? | 16:13 |
| Razique | why was I able to have two vlan interfaces without issue | 16:13 |
| *** freeflying has quit IRC | 16:13 | |
| Razique | while with Diablo I need a nova-network per nova-compute | 16:13 |
| Razique | two vlan interfaces = --vlan_interface that was not the same on every compute node | 16:14 |
| Razique | coli_: the private interfaces is on the same range | 16:15 |
| mjfork | Razique: i just got a single machine running nova-network with VLANs and many compute nodes, is this what you are trying to do | 16:15 |
| *** misheska has joined #openstack | 16:15 | |
| Razique | mjfork: ok so the issue here is that with cactus, I was able to have one nova-network with it's vlan_interface=eth0 | 16:15 |
| mjfork | oh man, no idea on cactus | 16:15 |
| Razique | and on every node vlan_interface=eth1, eth2 etc... | 16:16 |
| coli_ | razique: then you don't have a problem ;-) if it was routed via public interface then it would have been snated and 'cause nova-api provides meta-data based on source ip it wouldn't find it ;-( | 16:16 |
| *** primeministerp has joined #openstack | 16:16 | |
| Razique | coli_: yah I see what you mean | 16:16 |
| Razique | but my vlan issue remain weird though | 16:16 |
| Razique | :/ | 16:16 |
| *** guigui has quit IRC | 16:17 | |
| *** f-yang has joined #openstack | 16:17 | |
| *** fred_ has joined #openstack | 16:17 | |
| *** primeministerp has quit IRC | 16:17 | |
| Razique | mjfork: i'll do extra test | 16:17 |
| *** fred_ is now known as Guest6998 | 16:17 | |
| *** Guest6998 has left #openstack | 16:18 | |
| Razique | eg remove nova-network, for instance | 16:18 |
| Razique | and see how routring works | 16:18 |
| *** reed has quit IRC | 16:18 | |
| *** rnirmal has joined #openstack | 16:18 | |
| Razique | erf I just saw the isntance can't reach the metadata server now | 16:19 |
| Razique | ok i'll go home and debuf that | 16:19 |
| Razique | debug* | 16:19 |
| Razique | laters guys | 16:19 |
| *** Razique has quit IRC | 16:19 | |
| *** axisys has joined #openstack | 16:20 | |
| *** yshh has quit IRC | 16:21 | |
| *** adiantum has quit IRC | 16:24 | |
| *** debo-os has joined #openstack | 16:25 | |
| *** spiffxp has joined #openstack | 16:25 | |
| *** TheOsprey has quit IRC | 16:25 | |
| *** martines has quit IRC | 16:26 | |
| *** jaypipes has joined #openstack | 16:31 | |
| *** imsplitbit has joined #openstack | 16:32 | |
| *** juddm has joined #openstack | 16:36 | |
| *** pradeep has quit IRC | 16:37 | |
| *** dobber has quit IRC | 16:38 | |
| *** clopez has quit IRC | 16:38 | |
| *** oubiwann has quit IRC | 16:40 | |
| *** martines has joined #openstack | 16:41 | |
| *** oubiwann has joined #openstack | 16:43 | |
| *** livemoon has joined #openstack | 16:43 | |
| *** wilmoore has joined #openstack | 16:44 | |
| *** maplebed has joined #openstack | 16:46 | |
| *** katkee has joined #openstack | 16:47 | |
| *** misheska has quit IRC | 16:49 | |
| *** koolhead17 has joined #openstack | 16:53 | |
| *** dolphm has quit IRC | 16:53 | |
| *** marcuz has quit IRC | 16:54 | |
| *** clopez has joined #openstack | 16:54 | |
| *** negronjl has quit IRC | 16:59 | |
| *** misheska has joined #openstack | 16:59 | |
| *** misheska has quit IRC | 16:59 | |
| *** negronjl has joined #openstack | 16:59 | |
| *** bencherian has quit IRC | 17:01 | |
| *** anotherjesse has joined #openstack | 17:01 | |
| *** Ramonster has quit IRC | 17:03 | |
| *** nerens has quit IRC | 17:03 | |
| *** vladimir3p has joined #openstack | 17:04 | |
| *** Razique has joined #openstack | 17:04 | |
| Razique | Ping from the apple store | 17:04 |
| Razique | djee what are theses places =D | 17:04 |
| Razique | mjfork: still around ? | 17:04 |
| *** foexle has quit IRC | 17:05 | |
| *** ShawnB has joined #openstack | 17:05 | |
| zykes- | notmyname: around still ? | 17:05 |
| *** ShawnB has left #openstack | 17:06 | |
| notmyname | zykes-: yes | 17:07 |
| mjfork | Razique: yes | 17:08 |
| Razique | mjfork: are you using diablo ? | 17:08 |
| mjfork | yes | 17:09 |
| zykes- | notmyname: has anyone run swift on solaris ? | 17:09 |
| notmyname | zykes-: nexenta has mentioned some stuff on the mailing list (Caitlin Bestler). I'm not sure if anyone from there is here in IRC | 17:10 |
| Razique | mjfork: so you have several nodes but only one nova network installed ? | 17:10 |
| Razique | Still trying to figure out how the vlan_interface is used | 17:10 |
| mjfork | yes | 17:10 |
| mjfork | VLAN interface is the interface on node used to establish the VLAN | 17:11 |
| Razique | do you have a different vlan_interface per node ? | 17:11 |
| Razique | yah I know :D | 17:11 |
| mjfork | no | 17:12 |
| *** juddm has quit IRC | 17:12 | |
| mjfork | all of my vlans need to be created on eth1 | 17:12 |
| Razique | mjfork: ah ok | 17:12 |
| Razique | :( | 17:12 |
| mjfork | my cloud controlelr creates all VLANs on eth1 | 17:12 |
| mjfork | that is the network device with access to all ym compute nodes | 17:12 |
| Razique | yayah, that's why it's working | 17:12 |
| mjfork | what do you have? | 17:13 |
| Razique | yah* | 17:13 |
| Razique | I've different vlan_interface per node :D | 17:13 |
| Razique | and that is why without multi_host mode, the vlan cannot be created | 17:13 |
| *** llang629 has joined #openstack | 17:14 | |
| mjfork | so each node has a dedicated interface on teh cconteroller? | 17:14 |
| Razique | and it doesn't seem there have been changes on that part except the multi_host stuff | 17:14 |
| Razique | mjfork: yah, but it's never the same | 17:14 |
| Razique | on the nova-network, it's eth0 | 17:14 |
| Razique | eth1 sorry | 17:15 |
| Razique | while it's eth2 on all nodes | 17:15 |
| mjfork | well, just set vlan_interface on each of the nodes to eth2 | 17:15 |
| mjfork | that is used by nova-compute to setup vlans | 17:16 |
| *** Corollary has joined #openstack | 17:16 | |
| mjfork | on nova-network it would be eth0 | 17:16 |
| mjfork | etc | 17:17 |
| Razique | mjfork: that is what I use to have with Cactus :) | 17:17 |
| Razique | it wa working fine | 17:17 |
| Corollary | Hi I'm running a XenServer pool right now (non openstack) with 80 nodes, however XS technically only supports pools of size 16 for shared storage and live migration.. does OpenStack support larger pools for VM migration and shared storage? | 17:17 |
| Razique | but since I migrated to Diablo, it's no longer working ; I think it has something to do with the multi_host moe | 17:17 |
| Razique | mode | 17:17 |
| *** katkee has quit IRC | 17:18 | |
| Razique | the only solution I quicklly tried was to instal nova-network on the node, so it uses the vlan_interface flag | 17:18 |
| *** TheOsprey has joined #openstack | 17:18 | |
| *** llang629 has left #openstack | 17:19 | |
| mjfork | looking at mine | 17:20 |
| Razique | https://bugs.launchpad.net/nova/+bug/810563 | 17:20 |
| *** swill has joined #openstack | 17:20 | |
| Razique | it looks like the bridge is used after as Vish statew | 17:21 |
| Razique | states | 17:21 |
| mjfork | my nova-compute log says vlan_interface eth1 | 17:21 |
| mjfork | and looks to be creating the VLAN on that itnerface | 17:21 |
| Razique | mjfork: yup | 17:22 |
| Razique | mjfork: is it production ? | 17:23 |
| mjfork | define production? | 17:23 |
| Razique | I mean is your architecture here is production or your lab ? | 17:24 |
| *** cleverdevil has joined #openstack | 17:24 | |
| mjfork | it is in my lab for playing | 17:24 |
| mjfork | not a production setup. | 17:24 |
| Razique | would you try something for me ? | 17:24 |
| Razique | enable multi_host mode and see if it's breaks running network | 17:24 |
| mjfork | i tired it | 17:24 |
| Razique | (while you have running instances) | 17:24 |
| mjfork | it did | 17:25 |
| mjfork | i had all kinds of problems | 17:25 |
| *** CaptTofu has joined #openstack | 17:25 | |
| mjfork | my original try had multi_host with nova-network on every node | 17:25 |
| mjfork | i couldn't get it working quite right | 17:25 |
| mjfork | so i reverted back | 17:25 |
| Razique | mjfork: have u seen that link http://blog.defunct.ca/2011/09/19/using-nova-networks-multi_host-to-remove-spof/ | 17:25 |
| mjfork | yes | 17:25 |
| Razique | ok | 17:25 |
| mjfork | that is waht i tried | 17:25 |
| Razique | thanks for that feedback then | 17:25 |
| Razique | that means I'm screwed if the nodes reboot =D | 17:26 |
| mjfork | you must be doing production? | 17:26 |
| Razique | i did migration on production without multi_host mode | 17:26 |
| Razique | running fine | 17:26 |
| Razique | on my lab, I rebooted and that how I figured the issue | 17:26 |
| Razique | maybe If I can figure how to live enable multi_host mode i'm safe | 17:27 |
| *** PotHix has quit IRC | 17:27 | |
| Razique | but based and your feedback, and my brief tests, it won't be that easy hehe | 17:27 |
| *** nacx has quit IRC | 17:28 | |
| *** cleverdevil has left #openstack | 17:28 | |
| *** juddm has joined #openstack | 17:29 | |
| Razique | grrr that Apple Store makes me sick =D | 17:29 |
| Razique | everything is so polished and shiny | 17:29 |
| zykes- | :p | 17:29 |
| *** helfrez has quit IRC | 17:29 | |
| *** clauden has quit IRC | 17:30 | |
| mjfork | yes, i had issues, i would find someplace to paly with | 17:30 |
| *** clauden has joined #openstack | 17:30 | |
| Razique | mjfork: here is what 'ill try : 1- remove nova-network and multi_host from the node | 17:30 |
| Razique | since the vlan and the bridge won't be down | 17:30 |
| Razique | i'll be able to keep the instances runninn without routing issue | 17:31 |
| Razique | then 'ill reinstall nova-network and enable multi_host and see where it breaks the network | 17:31 |
| Kiall | the issue is probably going to the the default gateway inside the instances.. | 17:32 |
| Kiall | they will all point at 1 IP.. the wrong IP | 17:32 |
| swill | i am writing an auth middleware for swift right now and I am confused about something. according to the documentation (http://docs.openstack.org/bexar/openstack-object-storage/developer/content/ch02s01.html), there should be a first request with User and Key and then they will be given a token. How is the auth middleware supposed to handle this? none of the examples I have seen show that. | 17:32 |
| Kiall | wormman mentioned he made some edits so all nova-network nodes used the same fixed_ip.. | 17:33 |
| Kiall | No idea how though | 17:33 |
| Razique | Kiall: I could add a pre routing rule then ? | 17:33 |
| Razique | for the "transition" phase | 17:33 |
| notmyname | swill: the auth middleware doesn't handle the token part (well, it could, like swauth). the middleware is responsible for checking to see if a token if valid for the given request | 17:33 |
| swill | notmyname: who creates the token then? it has to be my middleware if i am integrating with cloudstack because they will not give a token otherwise. | 17:35 |
| *** reed has joined #openstack | 17:35 | |
| *** arBmind has joined #openstack | 17:36 | |
| swill | notmyname: so should I just handle the request to /v1.0 differently and create a token based on the X-Auth-User and X-Auth-Key? | 17:36 |
| notmyname | swill: your auth server. the middleware piece is the glue between the auth server and the proxy server. your auth middleware /could/ do everything (like swauth does https://github.com/gholt/swauth), but it's really just there to glue the parts together | 17:36 |
| notmyname | swill: exactly | 17:36 |
| * notmyname lunch time | 17:36 | |
| Razique | Kiall: mjfork http://osdir.com/ml/openstack-cloud-computing/2011-09/msg00283.html | 17:36 |
| swill | notmyname: ok, i will see what i can do. | 17:37 |
| mjfork | yes, saw that. | 17:37 |
| Kiall | Razique: what about it? | 17:37 |
| mjfork | and running nova_network + nova_api will alllow 169.254.169.254 to work | 17:37 |
| mjfork | which is what that specific things was aimed at | 17:37 |
| *** yehudasa has joined #openstack | 17:38 | |
| yehudasa | vishy: did you see our changes to the https://blueprints.launchpad.net/nova/+spec/ceph-s3-gateway blueprint, it was pretty much rewritten last week? | 17:38 |
| swill | notmyname: it looks like tempauth also handles this piece as well. i will review those two and see what i can come up with. | 17:39 |
| Razique | mjfork: ow ok ; so the Ha also concurs nova-api ? | 17:39 |
| Razique | if so that would be the solution thenĆØ | 17:39 |
| Razique | then* | 17:39 |
| mjfork | Razique: what do you mean? | 17:40 |
| *** Ryan_Lane|away has quit IRC | 17:40 | |
| *** dysinger has joined #openstack | 17:40 | |
| *** MarcMorata has joined #openstack | 17:40 | |
| mjfork | "so the Ha also concurs nova-api ?" | 17:41 |
| Razique | well in fact setting up multi_host mode seems to break metadata routing | 17:43 |
| Razique | setting up a local nova-api would resolve the issue maybe | 17:43 |
| *** andrewbogott has joined #openstack | 17:44 | |
| Razique | i'll try that on the lab tonight ; and I'll give you a feedback. is having two nova-api could be dangerous within the same zone ? | 17:45 |
| *** dgags has quit IRC | 17:45 | |
| *** dotdevops has joined #openstack | 17:46 | |
| *** rustam_ has quit IRC | 17:47 | |
| *** vdo has quit IRC | 17:47 | |
| *** rustam has joined #openstack | 17:48 | |
| *** vipul has joined #openstack | 17:48 | |
| Razique | gtg thanks mjfork | 17:49 |
| Razique | later | 17:49 |
| *** Razique has quit IRC | 17:49 | |
| *** rustam_ has joined #openstack | 17:52 | |
| *** nati2 has joined #openstack | 17:53 | |
| *** rustam has quit IRC | 17:55 | |
| *** rustam_ has quit IRC | 18:01 | |
| *** cctrieloff has quit IRC | 18:02 | |
| *** nerens has joined #openstack | 18:03 | |
| *** rods has quit IRC | 18:03 | |
| *** nerens has quit IRC | 18:04 | |
| *** AlanClark has quit IRC | 18:06 | |
| *** GheRivero has quit IRC | 18:08 | |
| *** jdurgin has joined #openstack | 18:10 | |
| *** livemoon has quit IRC | 18:11 | |
| uvirtbot | New bug: #899299 in nova "nova-manage expext to register images with int id" [Undecided,New] https://launchpad.net/bugs/899299 | 18:11 |
| uvirtbot | New bug: #899302 in nova "Context goes None intermittently due to calls to get_admin_context method" [Undecided,New] https://launchpad.net/bugs/899302 | 18:11 |
| uvirtbot | New bug: #899298 in openstack-qa "Can not get tenant quota from /{version}/{tenantID}/os-quota-sets" [Undecided,New] https://launchpad.net/bugs/899298 | 18:13 |
| *** BasTichelaar has quit IRC | 18:13 | |
| *** AlanClark has joined #openstack | 18:14 | |
| *** dysinger has quit IRC | 18:14 | |
| *** gyee has joined #openstack | 18:15 | |
| *** juddm has quit IRC | 18:16 | |
| *** bencherian has joined #openstack | 18:17 | |
| *** AaronSchulz has joined #openstack | 18:18 | |
| *** juddm has joined #openstack | 18:21 | |
| *** dgags has joined #openstack | 18:22 | |
| *** clopez has quit IRC | 18:25 | |
| *** jog0 has joined #openstack | 18:25 | |
| *** jog0 has left #openstack | 18:25 | |
| *** tmichael has quit IRC | 18:26 | |
| *** MarcMorata has quit IRC | 18:27 | |
| *** crescendo has joined #openstack | 18:27 | |
| *** dysinger has joined #openstack | 18:28 | |
| *** bencherian has quit IRC | 18:28 | |
| *** Corollary has quit IRC | 18:29 | |
| *** AlanClark has quit IRC | 18:29 | |
| *** dysinger has quit IRC | 18:29 | |
| *** woleium has joined #openstack | 18:31 | |
| uvirtbot | New bug: #899310 in tempest "Self parameter in rand_name(data_utils.py) is not needed" [Undecided,Invalid] https://launchpad.net/bugs/899310 | 18:32 |
| *** Ryan_Lane has joined #openstack | 18:34 | |
| *** lloydde has quit IRC | 18:34 | |
| *** ollie1 has left #openstack | 18:40 | |
| *** cdub has joined #openstack | 18:42 | |
| *** dragondm has quit IRC | 18:45 | |
| *** tmichael has joined #openstack | 18:46 | |
| *** bencherian has joined #openstack | 18:52 | |
| *** danishman has quit IRC | 18:53 | |
| vicelow | Should I use special hardware in order to use Openstack? | 18:57 |
| *** bengrue has joined #openstack | 18:58 | |
| *** nerens has joined #openstack | 18:58 | |
| Kiall | vicelow: that really depends on your needs.. | 18:59 |
| vicelow | thanks Kiall, | 19:00 |
| Kiall | But - Basically anything that can run VMs will work.. | 19:00 |
| Kiall | (ie a VT enabled CPU, and lots of RAM) | 19:00 |
| vicelow | Interesting!... I will read more about this | 19:01 |
| vicelow | thnks | 19:01 |
| *** dolphm has joined #openstack | 19:04 | |
| notmyname | swill: ya, tempauth does too | 19:05 |
| swill | notmyname: ya, i think i understand how it works a little better now. i will implement it as i think it should be done and i am sure the community will review my implementation once i have somethign working. | 19:06 |
| notmyname | swill: cloudstack is citrix, right? | 19:07 |
| *** tmichael has quit IRC | 19:07 | |
| swill | notmyname: it is now, they recently bought it. i believe it was an opensource project before they bought it. | 19:08 |
| swill | notmyname: it is still opensource, but citrix has a dev team working on it a lot right now. | 19:08 |
| *** dgags has quit IRC | 19:13 | |
| *** bencherian has quit IRC | 19:15 | |
| *** MarkAtwood has joined #openstack | 19:16 | |
| *** anotherjesse has quit IRC | 19:17 | |
| uvirtbot | New bug: #899330 in glance "create WADL for image-api" [Undecided,New] https://launchpad.net/bugs/899330 | 19:17 |
| *** reed has quit IRC | 19:18 | |
| *** zigo has quit IRC | 19:19 | |
| *** dysinger has joined #openstack | 19:21 | |
| *** dpippenger has joined #openstack | 19:25 | |
| *** bcwaldon has joined #openstack | 19:26 | |
| notmyname | gnu111: nice tweet :-) | 19:27 |
| gnu111 | notmyname: thanks!. I have been thinking about some use cases for swift. storing old log files we rarely use is a serious headache. | 19:28 |
| juddm | gnu111: but why would you need 3+ replicas of those files? | 19:29 |
| notmyname | gnu111: we use swift for that. we use the swift clusters to store logs that the swift clusters generate. it works really nicely | 19:29 |
| py | is it keystone part of diablo itself or plugin? | 19:29 |
| notmyname | py: keystone is a separate openstack project, but it's only officially a core project after diablo | 19:29 |
| gnu111 | juddm: I know. I had the same thought. i probably don't need all the replicas. | 19:29 |
| *** tmichael has joined #openstack | 19:30 | |
| notmyname | (although there are backports and dependencies) | 19:30 |
| py | notmyname: aha tnx | 19:30 |
| py | is it still plain text for passwords in diablo notmyname ? | 19:30 |
| py | like it was in B | 19:30 |
| notmyname | py: you've now reached the limit of my keystone knowledge :-) | 19:30 |
| notmyname | py: you should ask dolphm | 19:31 |
| py | :) | 19:31 |
| notmyname | gnu111: are you looking for more use cases or just thinking about a few in some more detail? | 19:32 |
| *** reed has joined #openstack | 19:32 | |
| *** bcwaldon has quit IRC | 19:33 | |
| *** egant has joined #openstack | 19:33 | |
| gnu111 | notmyname: Looking for use cases that could be useful for us here. Most of our users (academic researchers) are used to Lustre type systems for their data. So I am just trying to think how they could take advantage of swift. | 19:34 |
| zykes- | gnu111: store static files | 19:40 |
| zykes- | for one | 19:40 |
| *** mwhooker has joined #openstack | 19:46 | |
| *** reed has quit IRC | 19:50 | |
| *** afmtab has joined #openstack | 19:51 | |
| *** nati2 has quit IRC | 19:58 | |
| *** javiF has quit IRC | 20:00 | |
| juddm | gnu111: massive file share | 20:00 |
| *** reed has joined #openstack | 20:02 | |
| vicelow | Is node.js supported in OpenStack object storage in the web service API ? | 20:03 |
| *** AlanClark has joined #openstack | 20:03 | |
| *** rnorwood1 has joined #openstack | 20:05 | |
| *** rnorwood has quit IRC | 20:06 | |
| dolphm | py: passwords salting & hashing was added at the last minute in diablo (and is available in stable/diablo) ... we use passlib, fyi | 20:09 |
| zykes- | vicelow: what you mean by supported ? | 20:09 |
| *** GheRivero has joined #openstack | 20:10 | |
| *** bcwaldon has joined #openstack | 20:12 | |
| *** codepoet has joined #openstack | 20:13 | |
| vicelow | I just read that the currently supported API bindings for object storage are php, python, java, c# and ruby | 20:14 |
| *** sloop has quit IRC | 20:16 | |
| *** sloop has joined #openstack | 20:17 | |
| notmyname | vicelow: those are the language bindings that are for rackspace cloud files (which is a superset of swift--ie it includes rackspace-specific cdn features) | 20:17 |
| *** ziyadb has joined #openstack | 20:17 | |
| codepoet | I've been trying to get a test install of openstack up following the starter guide, but i'm having trouble actually starting an instance... everything looks okay but when i try to start an instance it sticks at pending, and nothing ever gets placed into /var/lib/nova/instances ... can anyone point me to where i need to look for the problem? | 20:18 |
| jaypipes | soren: ping | 20:20 |
| vicelow | notmyname: but coud I develop a node.js app that can get files from swift storage? | 20:20 |
| jaypipes | soren: could you make me a driver/admin for the tempest project please? I need to create milestones/series and assign bugs and stuff.. | 20:20 |
| notmyname | vicelow: I suppose so. the API for swift is HTTP REST. the language bindings simply wrap up that language's http libs and provide some helper functions. if you have the ability to do arbitrary requests and set headers on them in node.js, then you can easily make a node.js binding | 20:24 |
| *** dragondm has joined #openstack | 20:24 | |
| *** dragondm has quit IRC | 20:25 | |
| vicelow | thanks notmyname | 20:25 |
| *** renuka has joined #openstack | 20:25 | |
| *** dragondm has joined #openstack | 20:25 | |
| *** renuka has quit IRC | 20:26 | |
| *** afmtab has quit IRC | 20:27 | |
| *** bencherian has joined #openstack | 20:29 | |
| *** bencherian has quit IRC | 20:30 | |
| *** debo-os has quit IRC | 20:33 | |
| *** AlanClark has quit IRC | 20:34 | |
| *** rnorwood1 has quit IRC | 20:34 | |
| *** ameade has quit IRC | 20:34 | |
| *** ameade has joined #openstack | 20:34 | |
| *** rnorwood has joined #openstack | 20:34 | |
| *** localhost has quit IRC | 20:36 | |
| uvirtbot | New bug: #899374 in tempest "NameError: global name 'exceptions' is not defined when build times out" [Low,Triaged] https://launchpad.net/bugs/899374 | 20:36 |
| *** localhost has joined #openstack | 20:37 | |
| *** arBmind has quit IRC | 20:37 | |
| Glacee | notmyname: What do you think of this setup? 4U server nodes.. with 36 3TB 7.2k drives.. instead of using head node+ 45 bay JBOD? | 20:38 |
| *** mahmoh has quit IRC | 20:38 | |
| Glacee | If my calculations are right.. it gives around 9disk per U.. vs 10 with head node + 2 servers.. but I think at smaller case it would make more sense to use the 4U.. it should not take a big hit if a node goed down | 20:40 |
| Glacee | head node + 2 jbods* | 20:40 |
| *** bencherian has joined #openstack | 20:43 | |
| *** dysinger has quit IRC | 20:44 | |
| notmyname | Glacee: ya, it depends also on your needs (DC power availability, etc). but that would work find | 20:45 |
| uvirtbot | New bug: #899383 in tempest "Out of the box, tempest does not run cleanly, needs config file" [Medium,In progress] https://launchpad.net/bugs/899383 | 20:45 |
| Glacee | notmyname: ok thanks | 20:46 |
| *** dprince has quit IRC | 21:10 | |
| *** bcwaldon has quit IRC | 21:11 | |
| *** bcwaldon has joined #openstack | 21:11 | |
| *** lorin1 has quit IRC | 21:11 | |
| *** markvoelker has quit IRC | 21:11 | |
| *** rnorwood has quit IRC | 21:11 | |
| *** sniperd has joined #openstack | 21:12 | |
| sniperd | Hi All, having trouble getting over setting up permissions on swift - Im getting an error "ERROR: swift.cache could not be found in env!", Seems to think I do not have memcache setup -- however I do. Full error is here https://gist.github.com/e298cb0cebe9ca95bfbe | 21:13 |
| *** Sander^work has quit IRC | 21:13 | |
| sniperd | search only returns setting "allow_account_management = true" in config but its already set | 21:14 |
| *** Sander^work has joined #openstack | 21:14 | |
| *** rnorwood has joined #openstack | 21:14 | |
| sniperd | swift 1.2.0 installed from PPA ubuntu packages | 21:14 |
| *** sniperd has quit IRC | 21:15 | |
| *** sniperd has joined #openstack | 21:15 | |
| uvirtbot | New bug: #899396 in tempest "KeyError masks real errors in server tests" [Medium,In progress] https://launchpad.net/bugs/899396 | 21:16 |
| *** jeremyb has quit IRC | 21:18 | |
| *** doude has quit IRC | 21:18 | |
| *** jeremyb has joined #openstack | 21:18 | |
| vicelow | wich filesystem is recommended to use in a object server? | 21:19 |
| *** doude has joined #openstack | 21:19 | |
| *** TheOsprey has quit IRC | 21:20 | |
| *** sniperd has quit IRC | 21:21 | |
| *** sniperd has joined #openstack | 21:21 | |
| zykes- | vicelow: xfs | 21:22 |
| vicelow | thanks zykes- | 21:22 |
| *** nati2 has joined #openstack | 21:22 | |
| *** lorin1 has joined #openstack | 21:23 | |
| koolhead17 | zykes-: sir | 21:26 |
| koolhead17 | annegentle: around | 21:26 |
| *** gnu111 has quit IRC | 21:26 | |
| notmyname | sniperd: 2 things: 1) the cache error is probably because you don't have the cache middleware in the pipeline 2) 1.2.0 is pretty old. you should use a newer version. talk to mtaylor or ttx about the right PPA to pull it from | 21:27 |
| zykes- | koolhead17: ?:p | 21:27 |
| mtaylor | sniperd: what ppa are you installing from? | 21:28 |
| sniperd | notmyname: Ah... OK | 21:28 |
| koolhead17 | zykes-: how are you sirrrrrrrr | 21:28 |
| koolhead17 | hola mtaylor | 21:28 |
| mtaylor | hola koolhead17 | 21:28 |
| koolhead17 | mtaylor: am good. how have you been? | 21:28 |
| mtaylor | sniperd: try using ppa:openstack-release/2011.3 instead | 21:28 |
| mtaylor | koolhead17: excellent | 21:29 |
| koolhead17 | mtaylor: sounds great!! :) | 21:29 |
| sniperd | mtaylor: was using http://ppa.launchpad.net/swift-core/ppa/ubuntu | 21:29 |
| zykes- | koolhead17: good | 21:30 |
| koolhead17 | zykes-: nice | 21:30 |
| mtaylor | sniperd: that is very deprecated | 21:30 |
| mtaylor | notmyname: do you have an opposition to me deleting that ppa? | 21:30 |
| sniperd | mtaylor: that probably explains my problem(s) then, Thanks for the help, I appreciate it | 21:31 |
| mtaylor | sniperd: there is also a release ppa and a trunk ppa off of swift-core ... did you follow documentation recently which told you to use the one you're using? I'd like to fix it if possible | 21:31 |
| annegentle | koolhead17: hi, am now! | 21:31 |
| koolhead17 | annegentle: awesome | 21:32 |
| notmyname | mtaylor: I see no reason to keep the deprecated ppa around. I seriously doubt anyone is using it for prod (but I suppose it's possible), and I don't want them to use 1.2 anyway. | 21:32 |
| *** yamahata has quit IRC | 21:33 | |
| notmyname | sniperd: and to echo mtaylor, please let us know where you saw a reference to that old PPA. we need to make sure the docs are updated | 21:33 |
| sniperd | mtaylor: It's actually from a chef cookbook I found from voxel, Figured I would use that as a starting spot. https://github.com/voxeldotnet/openstack-swift-chef | 21:33 |
| *** yamahata has joined #openstack | 21:34 | |
| sniperd | If I get this in good working order I will submit it to opscode so theres a semi-official one | 21:35 |
| mtaylor | sniperd: is there no swift coverage in openstack-chef? | 21:36 |
| mtaylor | nope. great | 21:36 |
| mtaylor | sniperd: if you get it working, would you submit it to openstack/openstack-chef/ | 21:37 |
| mtaylor | ? | 21:37 |
| sniperd | mtaylor: yeah - I definitely will | 21:37 |
| mtaylor | sniperd: https://github.com/openstack/openstack-chef --- code submission guidelines at http://wiki.openstack.org/GerritWorkflow | 21:37 |
| mtaylor | awesome | 21:37 |
| *** dysinger has joined #openstack | 21:38 | |
| sniperd | perfect, thanks again for the help | 21:38 |
| mtaylor | notmyname: done | 21:38 |
| notmyname | mtaylor: thanks | 21:38 |
| *** debo-os has joined #openstack | 21:40 | |
| *** nerens_ has joined #openstack | 21:41 | |
| *** nerens has quit IRC | 21:41 | |
| *** nerens_ is now known as nerens | 21:41 | |
| *** ziyadb has quit IRC | 21:43 | |
| *** ziyadb has joined #openstack | 21:44 | |
| *** joesavak has joined #openstack | 21:45 | |
| *** Vinsh_HP has quit IRC | 21:46 | |
| *** mdomsch has joined #openstack | 21:51 | |
| uvirtbot | New bug: #899408 in swift "whitespace and line length violations" [Undecided,New] https://launchpad.net/bugs/899408 | 21:55 |
| *** lorin1 has quit IRC | 21:57 | |
| *** joesavak has quit IRC | 21:57 | |
| *** jfluhmann has joined #openstack | 21:57 | |
| *** codepoet has quit IRC | 21:58 | |
| *** mgoldmann has quit IRC | 22:00 | |
| *** adam_g has joined #openstack | 22:00 | |
| *** lts has quit IRC | 22:01 | |
| *** lmh has quit IRC | 22:03 | |
| *** lmh has joined #openstack | 22:04 | |
| *** llang629 has joined #openstack | 22:05 | |
| *** bencherian has quit IRC | 22:05 | |
| *** rsampaio has quit IRC | 22:06 | |
| *** bencherian has joined #openstack | 22:07 | |
| *** reed has quit IRC | 22:11 | |
| *** Guest3171 has quit IRC | 22:11 | |
| *** dolphm has quit IRC | 22:12 | |
| *** llang629 has left #openstack | 22:14 | |
| juddm | sniperd: as the author of those voxel recipes, I'd like to encourge you to look at andi abes's chef recipes instead. :) | 22:15 |
| *** dragondm has quit IRC | 22:15 | |
| juddm | It's rolled into crowbar now, but they work pretty OK without crowbar. | 22:15 |
| *** vicelow has quit IRC | 22:15 | |
| adam_g | heya, is keystone.conf's default_store unreleated to the database configured in sql_connection? | 22:16 |
| *** juddm has quit IRC | 22:16 | |
| *** rsampaio has joined #openstack | 22:18 | |
| *** russellb has quit IRC | 22:19 | |
| *** russellb has joined #openstack | 22:19 | |
| *** reed has joined #openstack | 22:22 | |
| *** dolphm has joined #openstack | 22:25 | |
| uvirtbot | New bug: #899414 in nova "whitespace and line length violations" [Undecided,New] https://launchpad.net/bugs/899414 | 22:26 |
| *** dubsquared has quit IRC | 22:26 | |
| maplebed | I'm trying to throw out some log messages from within one of the wsgi modules in swift-proxy. can anybody offer advice on how to do that? | 22:28 |
| *** vladimir3p has quit IRC | 22:28 | |
| notmyname | maplebed: if you are using the standard get_logger call in swift, anything you print to stdout will get logged (INFO level) | 22:29 |
| *** vernhart has quit IRC | 22:29 | |
| maplebed | that's in swift.common.utils! let's see if I can get that to work for me. | 22:30 |
| maplebed | thanks. | 22:30 |
| *** Vinsh_HP has joined #openstack | 22:32 | |
| *** dolphm has quit IRC | 22:32 | |
| *** bcwaldon has quit IRC | 22:38 | |
| *** miclorb_ has joined #openstack | 22:39 | |
| maplebed | offhand, notmyname, do you know if I need to kick the swift-proxy on each change to my python or does it pick up changes live? | 22:39 |
| notmyname | maplebed: you'll need to reload the proxy so it sees the updated middleware. swift-init proxy reload will do a graceful close with no client interruption. swift-init proxy restart will terminate existing connections and restart the process | 22:40 |
| maplebed | sweet. | 22:40 |
| * AaronSchulz looks up CDN sharing | 22:43 | |
| notmyname | maplebed: obviously, use a different name other than "proxy" if you are doing middleware on the other servers... | 22:44 |
| *** lborda has quit IRC | 22:44 | |
| *** vernhart has joined #openstack | 22:44 | |
| maplebed | sure. | 22:44 |
| notmyname | AaronSchulz: what is CDN sharing? | 22:45 |
| AaronSchulz | http://docs.openstack.org/bexar/openstack-object-storage/developer/content/ch04s01.html | 22:45 |
| notmyname | AaronSchulz: heh. that should be there | 22:46 |
| notmyname | annegentle: ^^ looks like there is a holdover from the cloud files docs | 22:46 |
| sniperd | juddm: will do :) | 22:47 |
| maplebed | notmyname: you mean "that shouldn't be there", right? | 22:48 |
| notmyname | AaronSchulz: that's also old docs | 22:48 |
| notmyname | maplebed: oh yeah | 22:48 |
| notmyname | AaronSchulz: heh. that shouldn't be there | 22:48 |
| AaronSchulz | I saw make_private() and is_public in cloudfiles.php when I was looking for access controls | 22:48 |
| *** ryker has quit IRC | 22:48 | |
| notmyname | AaronSchulz: the cdn stuff (in those old docs) is a holdover from where those docs came from (the rackspace cloud files dev guide). swift doesn't include any cdn integration | 22:49 |
| AaronSchulz | ok | 22:49 |
| notmyname | AaronSchulz: it seems the current docs are correct, though | 22:49 |
| AaronSchulz | is there a way to restrict access of containers to certain users (without custom python code) | 22:49 |
| notmyname | AaronSchulz: http://docs.openstack.org/api/openstack-object-storage/1.0/content/ | 22:49 |
| notmyname | AaronSchulz: that's up to the auth server you use. | 22:50 |
| notmyname | AaronSchulz: example for tempauth (and swauth): http://programmerthoughts.com/openstack/swift-permissions/ | 22:50 |
| *** lloydde has joined #openstack | 22:52 | |
| *** kbringard has quit IRC | 22:56 | |
| AaronSchulz | maplebed: I'm thinking that the auth for rewritten old-style urls should use a different user than when MW writes/reads things from the proxy via cloudfiles | 22:56 |
| AaronSchulz | notmyname: thanks | 22:57 |
| maplebed | sure. that'd let us give read-only vs. read-write perms to the files. | 22:58 |
| AaronSchulz | containers like xx-wikipedia-images-deleted would be restricted | 22:58 |
| *** bencherian has quit IRC | 23:02 | |
| *** bencherian has joined #openstack | 23:08 | |
| *** jj0hns0n has quit IRC | 23:08 | |
| *** GheRivero has quit IRC | 23:08 | |
| *** gyee has quit IRC | 23:11 | |
| *** rustam has joined #openstack | 23:12 | |
| *** ahasenack has quit IRC | 23:16 | |
| *** RicardoSSP has joined #openstack | 23:17 | |
| *** MarkAtwood has quit IRC | 23:17 | |
| *** debo-os_ has joined #openstack | 23:17 | |
| *** miclorb_ has quit IRC | 23:18 | |
| *** rnorwood has quit IRC | 23:19 | |
| *** debo-os has quit IRC | 23:20 | |
| *** debo-os_ is now known as debo-os | 23:21 | |
| *** jakedahn has joined #openstack | 23:21 | |
| *** jdurgin has quit IRC | 23:26 | |
| *** debo-os has quit IRC | 23:27 | |
| *** jdurgin has joined #openstack | 23:28 | |
| *** dolphm has joined #openstack | 23:28 | |
| *** imsplitbit has quit IRC | 23:41 | |
| *** jfluhmann has quit IRC | 23:44 | |
| *** code_franco has quit IRC | 23:47 | |
| *** koolhead17 has left #openstack | 23:47 | |
| *** dolphm has quit IRC | 23:50 | |
| *** rnirmal has quit IRC | 23:55 | |
| *** ziyadb has quit IRC | 23:56 | |
| *** deshantm_laptop_ has joined #openstack | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!
