*** andrewbogott is now known as andrewbogott_afk | 00:02 | |
*** andrewbogott_afk is now known as andrewbogott | 00:06 | |
*** andrewbogott is now known as andrewbogott_afk | 00:07 | |
*** pk has joined #openstack-dns | 00:40 | |
*** pk has quit IRC | 00:50 | |
*** ericpeterson has joined #openstack-dns | 00:56 | |
*** ericpeterson has quit IRC | 01:00 | |
*** MentalRay has quit IRC | 01:01 | |
*** MentalRay_ has joined #openstack-dns | 01:01 | |
*** andrewbogott_afk is now known as andrewbogott | 01:03 | |
*** pk has joined #openstack-dns | 01:03 | |
*** andrewbogott is now known as andrewbogott_afk | 01:03 | |
*** pk has quit IRC | 01:13 | |
*** pk has joined #openstack-dns | 01:20 | |
*** MentalRay_ has quit IRC | 01:26 | |
*** mwagner_lap has quit IRC | 01:27 | |
*** ericpeterson has joined #openstack-dns | 01:31 | |
*** ericpeterson has quit IRC | 01:32 | |
*** ericpeterson has joined #openstack-dns | 01:33 | |
*** johnbelamaric has joined #openstack-dns | 01:34 | |
*** ericpeterson has quit IRC | 01:34 | |
*** stanzgy has joined #openstack-dns | 01:38 | |
*** mwagner_lap has joined #openstack-dns | 01:39 | |
*** ericpeterson has joined #openstack-dns | 02:01 | |
*** mwagner_lap has quit IRC | 02:06 | |
*** pk has quit IRC | 02:12 | |
*** mwagner_lap has joined #openstack-dns | 02:19 | |
*** johnbelamaric has quit IRC | 02:26 | |
*** pk has joined #openstack-dns | 02:47 | |
*** pk__ has joined #openstack-dns | 02:50 | |
*** pk has quit IRC | 02:50 | |
*** pk has joined #openstack-dns | 02:51 | |
*** pk__ has quit IRC | 02:51 | |
*** ericpeterson has quit IRC | 02:56 | |
*** km has quit IRC | 03:00 | |
*** MentalRay has joined #openstack-dns | 03:01 | |
*** ericpeterson has joined #openstack-dns | 03:06 | |
*** ericpeterson has quit IRC | 03:08 | |
*** GonZo2K has joined #openstack-dns | 03:41 | |
*** GonZo2K has joined #openstack-dns | 03:41 | |
*** MentalRay has quit IRC | 04:05 | |
*** MentalRay has joined #openstack-dns | 04:05 | |
*** pk has quit IRC | 04:16 | |
*** richm has quit IRC | 04:39 | |
*** pk has joined #openstack-dns | 05:17 | |
*** GonZo2K has quit IRC | 05:36 | |
*** pk has quit IRC | 05:50 | |
openstackgerrit | OpenStack Proposal Bot proposed openstack/designate: Imported Translations from Transifex https://review.openstack.org/169630 | 06:04 |
---|---|---|
*** MentalRay has quit IRC | 06:44 | |
*** fawadkhaliq has joined #openstack-dns | 06:53 | |
*** timfreund has quit IRC | 07:03 | |
*** fawadkhaliq has quit IRC | 07:08 | |
*** jordanP has joined #openstack-dns | 07:09 | |
*** fawadkhaliq has joined #openstack-dns | 07:10 | |
*** timfreund has joined #openstack-dns | 07:12 | |
*** fawadkhaliq is now known as fawadkhaliq_ | 07:15 | |
*** fawadkhaliq_ is now known as fawadkhaliq | 07:15 | |
*** fawadkhaliq has quit IRC | 07:16 | |
*** fawadkhaliq has joined #openstack-dns | 07:17 | |
*** timfreund has quit IRC | 07:20 | |
*** timfreund has joined #openstack-dns | 07:20 | |
*** fawadkhaliq has quit IRC | 07:21 | |
*** fawadkhaliq has joined #openstack-dns | 07:22 | |
*** kei_yama has quit IRC | 08:44 | |
openstackgerrit | Endre Karlson proposed openstack/designate: Fix support for PostgreSQL https://review.openstack.org/169662 | 09:06 |
*** untriaged-bot has joined #openstack-dns | 10:01 | |
untriaged-bot | Untriaged bugs so far: | 10:01 |
untriaged-bot | https://bugs.launchpad.net/designate/+bug/1433650 | 10:01 |
openstack | Launchpad bug 1433650 in Designate "Floating IP Reverse DNS API cannot query neutron using passed through credentials" [Medium,New] - Assigned to Endre Karlson (endre-karlson) | 10:01 |
untriaged-bot | https://bugs.launchpad.net/designate/+bug/1437699 | 10:01 |
openstack | Launchpad bug 1437699 in Designate "mDNS should Handle Bad File Descriptor" [Medium,New] | 10:02 |
*** untriaged-bot has quit IRC | 10:02 | |
*** stanzgy_ has joined #openstack-dns | 10:36 | |
*** stanzgy has quit IRC | 10:36 | |
*** stanzgy_ has quit IRC | 10:52 | |
*** mwagner_lap has quit IRC | 11:20 | |
*** mwagner_lap has joined #openstack-dns | 12:09 | |
*** richm has joined #openstack-dns | 12:13 | |
*** MentalRay has joined #openstack-dns | 12:14 | |
*** busterswt has joined #openstack-dns | 12:16 | |
*** ericpeterson has joined #openstack-dns | 12:19 | |
*** johnbelamaric has joined #openstack-dns | 12:21 | |
*** ericpeterson has quit IRC | 12:34 | |
*** MentalRay has quit IRC | 12:41 | |
*** johnbelamaric has quit IRC | 12:55 | |
*** nkinder has quit IRC | 13:17 | |
Kiall | timsim: ;; XFR size: 1670 records (messages 2, bytes 88567) | 13:23 |
Kiall | fixed :) Sorting tests etc out now | 13:23 |
-openstackstatus- NOTICE: gerrit has been restarted to restore event streaming. any change events missed by zuul (between 12:48 and 13:28 utc) will need to be rechecked or have new approval votes set | 13:28 | |
Kiall | mugsie: ^ ;) | 13:28 |
mugsie | -_- | 13:29 |
Kiall | explains the lack of events | 13:29 |
mugsie | yup | 13:29 |
Kiall | timsim: your 40k record dump you gave me.. | 13:29 |
Kiall | ;; Query time: 17500 msec | 13:29 |
Kiall | ;; SERVER: 127.0.0.1#5354(127.0.0.1) | 13:29 |
Kiall | ;; WHEN: Wed Apr 01 13:28:57 UTC 2015 | 13:29 |
Kiall | ;; XFR size: 40958 records (messages 14, bytes 890505) | 13:29 |
Kiall | 17.5 seconds .. which still sucks compare to pDNS/BIND - But too bad a start.. | 13:30 |
Kiall | (Also - That's on a crappy VM) | 13:32 |
Kiall | Same AXFR on my desktop == ;; Query time: 9340 msec .. So, still room for improvement.. | 13:38 |
jbratton | it takes me 4.5 seconds for a zone that size on our production nameservers where everything is directly connected to the same switch | 13:43 |
jbratton | so I'd say that is really good | 13:43 |
elarson | o/ jbratton | 13:44 |
jbratton | hey | 13:44 |
* elarson didn't know jbratton hangs out here :) | 13:44 | |
jbratton | I lurk and throw out useless bits of wisdom here and there | 13:44 |
Kiall | jbratton: 50% slower ain't bad for pure python ;) | 13:45 |
jbratton | that's what I'm saying, 9 seconds is great for that | 13:45 |
jbratton | don't know how much better you'll get short of IXFR one day | 13:45 |
Kiall | But.. I know we're doing daft things still like iterating the 45k records more times than we need to | 13:45 |
Kiall | ahu: about? I'm curious what a zone approx sized 40958 records (messages 14, bytes 890505) takes to AXFR on pDNS+MySQL? (I don't have a pDNS DB that large handy, hopeing you know off hand :)) | 13:49 |
jbratton | sorry, it appears that I lied.. apparently native BIND is faster than dig AXFR | 13:53 |
jbratton | Transfer completed: 30 messages, 43146 records, 1288480 bytes, 0.699 secs (1843319 bytes/sec) | 13:53 |
jbratton | probably uses edns to speed things up | 13:53 |
Kiall | jbratton: I was testing with dif, curious what the query looks like.. any chance of a pcap? | 13:55 |
jbratton | yeah, I think I can do that.. let me play with dig for a minute though and see if I can reproduce that speed | 13:56 |
jbratton | hah, way better.. it appears dig is including the time to display it all on the terminal | 13:57 |
jbratton | try dig +noall +stats ... | 13:57 |
jbratton | I got back a response of 135ms.. even faster than BIND is reporting | 13:57 |
*** fawadkhaliq has quit IRC | 13:59 | |
Kiall | Approx the same, ;; Query time: 7331 msec | 13:59 |
*** nkinder has joined #openstack-dns | 14:00 | |
*** MentalRay has joined #openstack-dns | 14:00 | |
Kiall | Really just need to do some profiling, the hotspots will show themselves pretty quick.. | 14:00 |
jbratton | interesting.. then I guess I was just killing my own performance by logging in to a remote server and having to wait for it all to make it to my terminal | 14:00 |
Kiall | I wouldn't have thought dig would include render times? | 14:03 |
jbratton | me either, but it must have been | 14:03 |
jbratton | root@ans1:~# dig +noall +stats -b 1.2.3.4 AXFR slicehost.net | 14:03 |
jbratton | ;; Query time: 132 msec | 14:04 |
jbratton | ;; SERVER: 67.192.177.49#53(67.192.177.49) | 14:04 |
jbratton | ;; WHEN: Wed Apr 01 14:01:24 UTC 2015 | 14:04 |
jbratton | ;; XFR size: 43146 records (messages 30, bytes 1288480) | 14:04 |
jbratton | and that's with 9.9.6-P1, if it matters | 14:05 |
jbratton | I know they have made some improvement over the years if you are using an older version of bind utils | 14:05 |
*** ericpeterson has joined #openstack-dns | 14:05 | |
Kiall | `dig -v` tells me - 9.9.5-3ubuntu0.2-Ubuntu | 14:05 |
jbratton | close enough | 14:08 |
Kiall | Yep, bind does have the advantage of keeping all this in memory.. vs say pDNS which loads from the DB like we do here (I guess it heavily caches..) | 14:09 |
jbratton | yeah, which is one of the reasons we've never been able to use a db backend even though I would much rather do that than have 1.7 million zone files | 14:09 |
Kiall | Theres also a C lib with python bindings I've been eyeing up to replace dnspython next cycle, but I've done little more than eyeball the docs so fa | 14:10 |
Kiall | r | 14:10 |
jbratton | just rewrite it all in C, problem solved :) | 14:11 |
Kiall | ekarlso rewrote up a early verson of designate-mdns service in go, it was much faster as you can expect ;) | 14:11 |
mugsie | i want to write it all in c++, for the craic / interest | 14:12 |
mugsie | and try to relearn C++ after years of not using it | 14:12 |
jbratton | I just have a general dislike of object-oriented languages, but it's because I learned C as my first programming language | 14:13 |
*** paul_glass has joined #openstack-dns | 14:13 | |
mugsie | so did I - but i ended up using structs in a distinctly OO style | 14:14 |
mugsie | drove lectures in Uni mental | 14:14 |
jbratton | I've caught myself doing the reverse, trying out OO languages but then it all looks procedural | 14:15 |
jbratton | when they forced me to learn Java in school | 14:15 |
mugsie | :) - yeah, i have seen that before alright | 14:17 |
Kiall | Yea, so 1 loop of the records is taking 3.9 seconds.. And the second loop, which includes rendering, is taking 2.8... Both can be combined | 14:21 |
Kiall | (rendering as in writing out the wireformat to a in-mem buffer) | 14:22 |
*** andrewbogott_afk is now known as andrewbogott | 14:23 | |
jbratton | the worst part is that we don't have any customer zones larger than a few thousand records, it's our internal groups that like to throw everything in to one huge zone | 14:23 |
Kiall | jbratton: lol - I know that EXACT feeling. | 14:24 |
Kiall | 45k is small compared to 1 particular zone we have :( | 14:24 |
jbratton | ouch, haha | 14:24 |
*** fawadkhaliq has joined #openstack-dns | 14:53 | |
*** johnbelamaric has joined #openstack-dns | 15:01 | |
openstackgerrit | Tim Simmons proposed openstack/designate: Fixup updates in the agent backend https://review.openstack.org/169793 | 15:04 |
*** johnbelamaric1 has joined #openstack-dns | 15:06 | |
ekarlso | Kiall: what zone is that ? :p | 15:07 |
Kiall | Probably shouldn't talk about it here ;) | 15:07 |
ekarlso | :p | 15:08 |
*** pk__ has joined #openstack-dns | 15:09 | |
*** johnbelamaric has quit IRC | 15:09 | |
timsim | ^^^ Is something needed for rc-1. But that would be changed a bit based on the conversation yesterday Kiall, hopefully before the final release. (Getting that notify logic into the backends) | 15:11 |
Kiall | timsim: not sure what ^^^ is referring to? :) | 15:11 |
timsim | Sorry, the code change: https://review.openstack.org/169793 | 15:12 |
Kiall | Ah.. I thought it was a question - "Is there something needed for rc1" vs "That's needed for rc1" .. Right ;) | 15:13 |
*** pk__ has quit IRC | 15:13 | |
Kiall | timsim: So, not quite sure I understand what ^ is actually doing>? | 15:17 |
timsim | As it stands right now, the Agent isn't actually getting a NOTIFY at any point in the process. (I should have realized that earlier) | 15:17 |
Kiall | Ohh, I see.. So this is a temp thing until we do the stuff we talked about yesterday? | 15:19 |
Kiall | Well - It might actually be the first part of it? Just for the 1 backend cls tho | 15:19 |
Kiall | LOL - Spotted this on our internal chat.. https://s3.amazonaws.com/uploads.hipchat.com/8522/1679739/CdH5YH6xnHwJd5C/upload.png | 15:20 |
timsim | Yep. That's right, I just wasn't sure if I could get that done before rc1. So i wanted to make sure that got up. | 15:20 |
timsim | The other change is a bit more complicated, have to change mdns method signatures and stuff. | 15:21 |
Kiall | Yep, makes sense | 15:21 |
timsim | LOL that thing is awesome! | 15:21 |
Kiall | I've wanted to change the PoolNameserver arg to mDNS into host:port ;) | 15:21 |
timsim | Yeah, I was just going down that road, realized it might be a bit more work, so I wanted to get this thing up quickly :) | 15:21 |
*** paul_glass has quit IRC | 15:22 | |
*** paul_glass has joined #openstack-dns | 15:23 | |
elarson | just did a rebase and my config broke. I'm fixing up the pool config and I'm wondering the target is vs. nameservers? | 15:47 |
*** jmcbride has joined #openstack-dns | 15:47 | |
Kiall | elarson: so, wondering about the concept or how to write the config? | 15:49 |
openstackgerrit | Graham Hayes proposed openstack/designate-specs: Open Liberty Specs https://review.openstack.org/169816 | 15:49 |
elarson | Kiall: just wanted clarification on the config. | 15:49 |
* elarson suspects we switched from "nameservers" to "targets", but isn't really sure | 15:49 | |
Kiall | So - nameservers got split into 2 .. The easist way to think about it is, for support providers like DynECT or Akamai, we write to one place (Create a new zone please sent to the Akamai API), while we query for status N places (ns1.akamai.com, ns2.akamai.com, etc) | 15:51 |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: Ensure mDNS can AXFR zones over 65k https://review.openstack.org/169818 | 15:51 |
elarson | ah ok, that makes sense | 15:51 |
Kiall | timsim / mugsie: ^ | 15:51 |
Kiall | config ends up a little like this: http://paste.openstack.org/show/197803/ | 15:53 |
Kiall | Humm.. Seem to get failures every other run on that patchset -_- | 16:00 |
elarson | Kiall: thanks! that is helpful | 16:01 |
Kiall | Oh cmon :( records coming out of the AXFR if different orders each time.. Grr | 16:01 |
*** untriaged-bot has joined #openstack-dns | 16:01 | |
untriaged-bot | Untriaged bugs so far: | 16:01 |
untriaged-bot | https://bugs.launchpad.net/designate/+bug/1433650 | 16:01 |
openstack | Launchpad bug 1433650 in Designate "Floating IP Reverse DNS API cannot query neutron using passed through credentials" [Medium,New] - Assigned to Endre Karlson (endre-karlson) | 16:01 |
Kiall | query = query.order_by(tables.recordsets.c.id) | 16:01 |
Kiall | ID's are UUIDs, which means.. BOO :( | 16:01 |
untriaged-bot | https://bugs.launchpad.net/designate/+bug/1437699 | 16:01 |
openstack | Launchpad bug 1437699 in Designate "mDNS should Handle Bad File Descriptor" [Medium,New] | 16:01 |
*** untriaged-bot has quit IRC | 16:02 | |
*** kfox1111 has joined #openstack-dns | 16:02 | |
kfox1111 | ok. trying to transfer a zone between tenants. | 16:03 |
kfox1111 | The request worked. now trying to accept it. | 16:03 |
kfox1111 | Its failing: {"code": 405, "type": "method_not_allowed", "request_id": "req-3f81a22f-114a-4e08-abbf-b9eaf9a2bfbe"} | 16:03 |
kfox1111 | any ideas? | 16:04 |
kfox1111 | using curl: curl -H "X-Auth-Token: $TOKEN" -H "Content-Type: application/json" -d @transfer2.json http://localhost:9001/v2/zones/tasks/transfer_accept -vvv | 16:04 |
timsim | I think you need to -X POST | 16:05 |
mugsie | kfox1111: yeah, it has to be a POST request | 16:06 |
*** MentalRay has quit IRC | 16:06 | |
*** ericpete_ has joined #openstack-dns | 16:07 | |
openstackgerrit | Merged openstack/designate: Fixup updates in the agent backend https://review.openstack.org/169793 | 16:07 |
*** ericpeterson has quit IRC | 16:08 | |
kfox1111 | -d defautls to post. | 16:09 |
kfox1111 | > POST /v2/zones/tasks/transfer_accept HTTP/1.1 | 16:09 |
mugsie | kfox1111: what is in the designate-api logs ? | 16:10 |
mugsie | i tested this yesterday - and it worked | 16:10 |
kfox1111 | not much... bunch of keystone auth stuff, and; 2015-04-01 16:12:49.217 6918 INFO eventlet.wsgi [req-9f23af95-afd6-4678-a9c8-568fdd03d105 ec31727f691e49dc71b08293dd340af9b12e48fd8285e9a2f2903e5dd79aea4a bf07b33d835e41e8b076843e291c1111 - - -] 127.0.0.1 - - [01/Apr/2015 16:12:49] "POST /v2/zones/tasks/transfer_accept HTTP/1.1" 405 292 0.391178 | 16:11 |
kfox1111 | This is from trunk from last thursday I think. | 16:12 |
kfox1111 | should I update to current? I was hoping to stay at a version that I would have a hope of using the v2 cli with until it is forward ported. | 16:13 |
mugsie | kfox1111: missing the 's' at the end ? | 16:13 |
mugsie | should be /v2/zones/tasks/transfer_accepts | 16:14 |
kfox1111 | on the end of what? transfer_accepts? | 16:14 |
kfox1111 | ah. let me see.. | 16:14 |
mugsie | that error code is wrong though -_- | 16:14 |
kfox1111 | that did it. Thanks. :) | 16:14 |
kfox1111 | I think maybe the docs are wrong then. | 16:15 |
mugsie | they could well be | 16:15 |
timsim | yeah, in the actual example http://docs.openstack.org/developer/designate/rest/v2/zones.html#accept-a-transfer-request | 16:15 |
kfox1111 | hmm... partially wrong. http://docs-draft.openstack.org/09/167809/5/check/gate-designate-docs/4ae5f17//doc/build/html/rest/v2/zones.html#post--zones-tasks-transfer_accepts | 16:15 |
kfox1111 | The example is missing the s. | 16:15 |
ekarlso | kfox1111: client has support for it too :p | 16:15 |
ekarlso | https://cloudistic.net/blog/designate-zone-transfers.html | 16:16 |
kfox1111 | http://specs.openstack.org/openstack/designate-specs/specs/juno/zone-migration-between-tenants.html also lacks the s. | 16:16 |
kfox1111 | ekarlso: I heard that the patches didn't currently work against trunk a couple of days ago. Did you fix it? | 16:17 |
kfox1111 | ekarlso: Awesome doc. thanks. :) | 16:18 |
ekarlso | kfox1111: nope, but about to fire it up :) | 16:19 |
kfox1111 | ok. cool. :) | 16:19 |
kfox1111 | arg... | 16:21 |
* kfox1111 doesn't understand why the cli doesn't support 'help' by itself... | 16:21 | |
kfox1111 | help with no args should be the same as -h. | 16:22 |
*** MentalRay has joined #openstack-dns | 16:23 | |
*** pk__ has joined #openstack-dns | 16:24 | |
kfox1111 | Message | Provided object does not match schema | 16:26 |
kfox1111 | any way to further figureout what part didn't match? | 16:26 |
ekarlso | kfox1111: you using the CLI ? | 16:26 |
kfox1111 | the older cli. 'designate record-create' | 16:27 |
mugsie | kfox1111: if you want to paste the object, I can tell you :) | 16:27 |
kfox1111 | still haven't patched keystone to work with the v2 api. | 16:27 |
kfox1111 | designate record-create --name kfox.pilot.cloud.pnnl.gov. --type NS --data 172.18.4.2 f02ac945-5d7e-4799-bfae-7a98f0c6f4c8 | 16:27 |
mugsie | NS have to be domain names | 16:28 |
mugsie | NS Records* | 16:28 |
mugsie | it should have said that tbh | 16:28 |
kfox1111 | k. | 16:28 |
* mugsie goes looking | 16:28 | |
*** pk__ has quit IRC | 16:29 | |
kfox1111 | hmm... ok. used a dns name... got a worse error. :/ | 16:29 |
kfox1111 | | Message | RecordSet belongs in a child zone: kfox.pilot.cloud.pnnl.gov. | | 16:29 |
ekarlso | :P | 16:29 |
kfox1111 | but the ns record for that should be in the parent, right? | 16:30 |
kfox1111 | kfox is a subdomain. I created and transfered to the project named "kfox". | 16:30 |
mugsie | kfox1111: in designate it doesnt have to - as they are on the same DNS server | 16:30 |
mugsie | so there would be 2 conflicting records | 16:31 |
kfox1111 | oh. so it automatically takes care of that? | 16:31 |
mugsie | those entries are auto generated by designate | 16:31 |
mugsie | yup | 16:31 |
kfox1111 | ok. cool. :) | 16:31 |
kfox1111 | even better. :) | 16:31 |
kfox1111 | ok. lets test this out... | 16:31 |
*** crc32 has joined #openstack-dns | 16:37 | |
*** crc32 has quit IRC | 16:38 | |
Kiall | mugsie / timsim - tests fixed to work every time ;) | 16:42 |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: Ensure mDNS can AXFR zones over 65k https://review.openstack.org/169818 | 16:42 |
*** jordanP has quit IRC | 16:51 | |
*** GonZo2K has joined #openstack-dns | 17:10 | |
*** GonZo2K has quit IRC | 17:14 | |
*** penick has joined #openstack-dns | 17:14 | |
*** ericpete_ has quit IRC | 17:24 | |
*** ericpeterson has joined #openstack-dns | 17:25 | |
*** pk__ has joined #openstack-dns | 17:33 | |
*** pk___ has joined #openstack-dns | 17:34 | |
*** pk__ has quit IRC | 17:34 | |
ahu | 15:49 <Kiall> ahu: about? I'm curious what a zone approx sized 40958 records (messages 14, bytes 890505) takes to AXFR on pDNS+MySQL? (I don't have a pDNS DB that large handy, hopeing you know off hand :)) | 17:42 |
ahu | do you mean in seconds? | 17:42 |
mugsie | seconds / milliseconds | 17:42 |
ahu | the actual time spent is dominated by the commit at the end | 17:43 |
ahu | for most backends | 17:43 |
ahu | which could be whole seconds | 17:43 |
ahu | but even if it is a tenth of a second, the commit will dominate | 17:43 |
Kiall | ahu: Ah, I meant serving an AXFR as a master rather than slaving off another server | 17:43 |
Kiall | e.g. read-only on the DB | 17:43 |
ahu | ah! - well, that should zoom, but I don't have any exact numbers | 17:44 |
Kiall | Fair enough :) I'll have to go test it out to find some numbers for comparison | 17:45 |
ahu | the main thing with axfr performance in our experience is if you stuff multiple records in one message | 17:49 |
ahu | otherwise it is a lot of work | 17:49 |
Kiall | Yep, we're currently stuffing packets until they hit ~65k bytes | 17:51 |
ahu | btw, such huge packets should not be compressed | 17:53 |
ahu | perhaps that is slowing you down | 17:54 |
Kiall | ahu: possibly, I'll look into that | 18:00 |
openstackgerrit | Graham Hayes proposed openstack/designate-specs: Add graded backends https://review.openstack.org/169859 | 18:00 |
ahu | compression is an O(N^2) affair | 18:00 |
Kiall | and - sadly - the lib were using doesn't give us a flag to disable name compression | 18:01 |
ahu | on a 64kbyte packet it kills you | 18:01 |
*** ericpeterson has quit IRC | 18:07 | |
*** GonZo2K has joined #openstack-dns | 18:09 | |
*** GonZo2K has joined #openstack-dns | 18:09 | |
Kiall | Seems to make no difference with this lib (I hacked it out ;)) quicky: | 18:10 |
Kiall | ;; Query time: 7425 msec | 18:10 |
Kiall | ;; XFR size: 40958 records (messages 14, bytes 890505) | 18:10 |
Kiall | ;; Query time: 7869 msec | 18:10 |
Kiall | ;; XFR size: 40958 records (messages 23, bytes 1505181) | 18:10 |
Kiall | Either way, I can see a easy win to get back ~50% of that time, which leaves us "good enough" for Kilo and can dig into C rather than python libs in the next cycle | 18:11 |
*** MentalRay has quit IRC | 18:18 | |
*** ericpeterson has joined #openstack-dns | 18:29 | |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: Update periodic sync defaults https://review.openstack.org/169872 | 18:31 |
Kiall | timsim / mugsie: changes discussed @ meet ^ | 18:32 |
Kiall | timsim: also - vinod away or something? | 18:32 |
timsim | Yep. He'll be back tomorrow | 18:33 |
timsim | Lol. Clippy is dancing around my review page | 18:34 |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: Correct min-ttl default config value https://review.openstack.org/169875 | 18:36 |
Kiall | timsim: Yep - Seems the openstack-infra folks have a sense of humor | 18:36 |
timsim | Should we port the "touch" functionality before release? | 18:36 |
Kiall | ehh, i seem to remember thinking it's not really necessary anymore with Pool Manager? But I can't remember what made me think that. | 18:37 |
timsim | Pool Manager should figure it out eventually, but I'd think there are times where you'd like to be able to get it before the next sync in 30 minutes. | 18:38 |
Kiall | True - Sometimes you just want to give a zone a "kick" | 18:38 |
timsim | Yep | 18:38 |
*** pk___ has quit IRC | 18:44 | |
*** andrewbogott is now known as andrewbogott_afk | 18:45 | |
*** pk__ has joined #openstack-dns | 18:46 | |
*** ericpeterson has quit IRC | 18:48 | |
*** jmcbride has quit IRC | 19:01 | |
*** ericpeterson has joined #openstack-dns | 19:02 | |
*** MentalRay has joined #openstack-dns | 19:05 | |
*** GonZo2K has quit IRC | 19:05 | |
*** jmcbride has joined #openstack-dns | 19:08 | |
kfox1111 | yay. teanant controlled subdomain. :) | 19:09 |
kfox1111 | working from any system at the lab. :) | 19:10 |
*** pk__ has quit IRC | 19:11 | |
*** pk__ has joined #openstack-dns | 19:14 | |
*** crc32 has joined #openstack-dns | 19:14 | |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: DevStack: Akamai/Dyn Support for Jenkins Credential Binding https://review.openstack.org/169891 | 19:18 |
ahu | 7 seconds for 40k records is excessive | 19:19 |
ahu | by any measure | 19:20 |
ahu | 'sudo perf top' usually tells a lot | 19:20 |
*** andrewbogott_afk is now known as andrewbogott | 19:31 | |
*** pk__ has quit IRC | 19:43 | |
kfox1111 | hmm... nova.openstack.common.notifier.rpc_notifier.... I'm not seeing the code for that. | 19:55 |
kfox1111 | what should it be set to? | 19:55 |
openstackgerrit | Merged openstack/designate: Update periodic sync defaults https://review.openstack.org/169872 | 20:01 |
*** busterswt has quit IRC | 20:07 | |
ekarlso | kfox1111: ? | 20:11 |
*** GonZo2K has joined #openstack-dns | 20:26 | |
*** ericpete_ has joined #openstack-dns | 20:30 | |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: DevStack: Akamai/Dyn Support for Jenkins Credentials and Exercices https://review.openstack.org/169891 | 20:31 |
*** ericpeterson has quit IRC | 20:32 | |
*** GonZo2K has quit IRC | 20:37 | |
*** fawadkhaliq has quit IRC | 20:39 | |
*** jmcbride has quit IRC | 20:46 | |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: DevStack: Akamai/Dyn Support for Jenkins Credentials and Exercises https://review.openstack.org/169891 | 20:48 |
*** pk__ has joined #openstack-dns | 20:50 | |
openstackgerrit | John Belamaric proposed openstack/designate: Infoblox backend for Designate https://review.openstack.org/138824 | 20:54 |
johnbelamaric1 | kiall: i got our backend end working a little bit - at least i can create/delete domains. but it looks like the create_record API never gets called in our backend. so, i am guessing the contract changed here | 20:58 |
*** jmcbride has joined #openstack-dns | 21:04 | |
*** nkinder has quit IRC | 21:11 | |
openstackgerrit | Merged openstack/designate: Ensure mDNS can AXFR zones over 65k https://review.openstack.org/169818 | 21:13 |
*** MentalRay has quit IRC | 21:18 | |
*** MentalRay has joined #openstack-dns | 21:19 | |
*** boris-42 has quit IRC | 21:28 | |
*** GonZo2K has joined #openstack-dns | 21:28 | |
*** GonZo2K has quit IRC | 21:30 | |
*** mwagner_lap has quit IRC | 21:36 | |
openstackgerrit | Merged openstack/designate: Correct min-ttl default config value https://review.openstack.org/169875 | 21:40 |
openstackgerrit | Merged openstack/designate: Akamai: mDNS must run on port 53 https://review.openstack.org/169479 | 21:43 |
*** jmcbride has quit IRC | 21:43 | |
openstackgerrit | Kiall Mac Innes proposed openstack/designate: Implement default page size for V2 API https://review.openstack.org/142505 | 21:51 |
*** pk__ has quit IRC | 21:51 | |
*** boris-42 has joined #openstack-dns | 21:57 | |
*** ekarlso has quit IRC | 22:04 | |
*** ekarlso has joined #openstack-dns | 22:04 | |
*** GonZo2K has joined #openstack-dns | 22:07 | |
*** GonZo2K has quit IRC | 22:09 | |
*** mwagner_lap has joined #openstack-dns | 22:14 | |
*** paul_glass has quit IRC | 22:14 | |
*** GonZo2K has joined #openstack-dns | 22:26 | |
*** nkinder has joined #openstack-dns | 22:34 | |
openstackgerrit | John Belamaric proposed openstack/designate: Infoblox backend for Designate https://review.openstack.org/138824 | 22:38 |
*** GonZo2K has quit IRC | 22:59 | |
*** pk has joined #openstack-dns | 23:02 | |
*** pk has quit IRC | 23:06 | |
*** chlong has joined #openstack-dns | 23:07 | |
*** penick has quit IRC | 23:11 | |
*** pk has joined #openstack-dns | 23:11 | |
*** jmcbride has joined #openstack-dns | 23:14 | |
*** kei_yama has joined #openstack-dns | 23:19 | |
*** ericpete_ has quit IRC | 23:23 | |
*** GonZo2K has joined #openstack-dns | 23:24 | |
*** GonZo2K has joined #openstack-dns | 23:24 | |
*** pk has quit IRC | 23:25 | |
*** pk has joined #openstack-dns | 23:28 | |
*** pk has quit IRC | 23:54 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!