Tuesday, 2016-02-02

« Monday, 2016-02-01 Index Wednesday, 2016-02-03 »
*** pdesai has quit IRC00:05
*** mp11 has quit IRC00:16
*** su_zhang has quit IRC00:21
*** su_zhang has joined #openstack-barbican00:21
*** ccneill has quit IRC00:22
*** rellerreller has joined #openstack-barbican00:26
*** dave-mcc_ has joined #openstack-barbican00:27
*** dave-mccowan has quit IRC00:28
*** mp1 has joined #openstack-barbican00:31
*** mp1 has quit IRC00:36
*** jamielennox|away is now known as jamielennox00:55
*** rellerreller has quit IRC00:59
*** jamielennox is now known as jamielennox|away01:23
*** cheneydc has joined #openstack-barbican01:29
*** dave-mcc_ has quit IRC02:09
*** dave-mccowan has joined #openstack-barbican02:10
*** su_zhang has quit IRC02:11
*** Nirupama has joined #openstack-barbican02:15
*** jhfeng has joined #openstack-barbican02:19
*** jhfeng has quit IRC02:20
*** jamielennox|away is now known as jamielennox02:44
*** zz_dimtruck is now known as dimtruck02:44
*** yuanying_ has joined #openstack-barbican03:21
*** yuanying has quit IRC03:24
*** yuanying has joined #openstack-barbican03:41
*** yuanying_ has quit IRC03:44
*** pdesai has joined #openstack-barbican03:47
*** yuanying has quit IRC04:05
*** yuanying has joined #openstack-barbican04:06
*** su_zhang has joined #openstack-barbican04:06
*** yuanying_ has joined #openstack-barbican04:07
*** yuanying has quit IRC04:07
*** kebray has joined #openstack-barbican04:09
*** jamielennox is now known as jamielennox|away04:10
*** kebray has quit IRC04:14
*** kebray_ has joined #openstack-barbican04:14
*** woodster_ has quit IRC04:16
*** sidx64 has joined #openstack-barbican04:25
*** sidx64_Cern has joined #openstack-barbican04:28
*** sidx64 has quit IRC04:31
*** jamielennox|away is now known as jamielennox04:39
*** fnaval has quit IRC04:47
*** sidx64_Cern is now known as sidx6404:53
*** pdesai has quit IRC05:09
*** dave-mccowan has quit IRC05:10
*** jamielennox is now known as jamielennox|away05:13
*** fnaval has joined #openstack-barbican05:16
*** cheneydc has quit IRC05:26
*** Nirupama has quit IRC05:32
*** gyee has quit IRC05:32
*** Nirupama has joined #openstack-barbican05:47
*** cheneydc has joined #openstack-barbican05:52
*** dimtruck is now known as zz_dimtruck06:00
*** alee has quit IRC06:06
openstackgerritFernando Diaz proposed openstack/castellan: Introduce Castellan Credential Objects  https://review.openstack.org/27060206:34
openstackgerritFernando Diaz proposed openstack/castellan: Introduce Castellan Credential Objects  https://review.openstack.org/27060206:47
*** fnaval has quit IRC06:52
*** fnaval has joined #openstack-barbican06:52
openstackgerritFernando Diaz proposed openstack/castellan: WIP: Introduce Castellan Credential Factory  https://review.openstack.org/27386306:58
openstackgerritFernando Diaz proposed openstack/castellan: WIP: Introduce Castellan Credential Factory  https://review.openstack.org/27386307:15
openstackgerritFernando Diaz proposed openstack/castellan: Allow Barbican Key Manager to accept different auth credentials  https://review.openstack.org/27387207:15
*** fnaval_ has joined #openstack-barbican07:17
*** fnaval has quit IRC07:20
openstackgerritFernando Diaz proposed openstack/castellan: Allow Barbican Key Manager to accept different auth credentials  https://review.openstack.org/27387207:31
*** scheuran has joined #openstack-barbican07:33
*** su_zhang has quit IRC07:52
*** kebray_ has quit IRC08:01
*** Nirupama has quit IRC08:25
*** Nirupama has joined #openstack-barbican08:41
*** cheneydc has quit IRC10:01
*** openstackgerrit has quit IRC10:17
*** openstackgerrit has joined #openstack-barbican10:17
*** dave-mccowan has joined #openstack-barbican10:59
*** sidx64_Cern has joined #openstack-barbican11:10
*** sidx64 has quit IRC11:14
*** dave-mccowan has quit IRC11:17
*** sid_cerner has joined #openstack-barbican11:48
*** sid_cerner is now known as sidx6411:48
*** sidx64_Cern has quit IRC11:51
*** sidx64 has quit IRC12:35
*** DuncanT has quit IRC12:36
*** ptoohill has quit IRC12:36
*** DuncanT has joined #openstack-barbican12:37
*** ptoohill has joined #openstack-barbican12:38
*** cheneydc has joined #openstack-barbican12:47
*** sidx64 has joined #openstack-barbican12:48
*** anteaya has joined #openstack-barbican13:00
*** anteaya has quit IRC13:06
*** rellerreller has joined #openstack-barbican13:39
*** su_zhang has joined #openstack-barbican13:41
*** Nirupama has quit IRC13:53
*** woodster_ has joined #openstack-barbican14:01
*** dave-mccowan has joined #openstack-barbican14:15
*** zz_dimtruck is now known as dimtruck14:18
*** edtubill has joined #openstack-barbican14:24
*** cheneydc has quit IRC14:26
*** kebray has joined #openstack-barbican14:30
*** dimtruck is now known as zz_dimtruck14:41
*** zz_dimtruck is now known as dimtruck14:42
*** cheneydc has joined #openstack-barbican14:52
*** dimtruck is now known as zz_dimtruck14:55
openstackgerritElvin Tubillara proposed openstack/barbican: Simple soft deletion cleanup script  https://review.openstack.org/26990315:00
*** jhfeng has joined #openstack-barbican15:00
*** rellerreller has quit IRC15:03
*** rellerreller has joined #openstack-barbican15:08
*** rellerreller has quit IRC15:09
*** rellerreller has joined #openstack-barbican15:10
*** zz_dimtruck is now known as dimtruck15:18
*** kebray has quit IRC15:21
*** jaosorior has joined #openstack-barbican15:23
*** kebray has joined #openstack-barbican15:37
*** mp1 has joined #openstack-barbican15:39
*** mp1 has quit IRC15:41
*** mp1 has joined #openstack-barbican15:44
*** mp1 has joined #openstack-barbican15:45
*** silos has joined #openstack-barbican15:48
*** mp1 has joined #openstack-barbican15:48
*** mp1 has quit IRC15:48
*** mp1 has joined #openstack-barbican15:49
*** kebray has quit IRC15:51
*** rellerreller has quit IRC15:55
*** kebray has joined #openstack-barbican15:56
*** sidx64 has quit IRC15:58
*** rellerreller has joined #openstack-barbican15:59
*** david-lyle has joined #openstack-barbican16:02
*** kebray has quit IRC16:05
*** kebray has joined #openstack-barbican16:06
*** su_zhang has quit IRC16:08
*** cheneydc has quit IRC16:08
*** su_zhang has joined #openstack-barbican16:13
*** ccneill has joined #openstack-barbican16:16
*** diazjf has joined #openstack-barbican16:16
*** diazjf has quit IRC16:16
*** diazjf has joined #openstack-barbican16:17
*** rtmorgan has joined #openstack-barbican16:19
*** woodster_ has quit IRC16:26
*** spotz_zzz is now known as spotz16:31
*** edtubill has quit IRC16:32
*** kebray has quit IRC16:36
*** pwp has joined #openstack-barbican16:41
*** fnaval_ has quit IRC16:48
*** jaosorior has quit IRC16:49
*** su_zhang has quit IRC16:52
*** gariveradlt has joined #openstack-barbican16:55
*** scheuran has quit IRC16:57
*** kebray has joined #openstack-barbican17:02
*** mp1 has quit IRC17:04
*** mp1 has joined #openstack-barbican17:05
*** silos has quit IRC17:06
*** fnaval has joined #openstack-barbican17:09
*** silos has joined #openstack-barbican17:12
*** pwp has quit IRC17:12
openstackgerritFernando Diaz proposed openstack/castellan: Add created property to Managed Objects  https://review.openstack.org/23815017:16
*** spotz is now known as spotz_zzz17:19
*** ccneill_ has joined #openstack-barbican17:31
*** openstackgerrit has quit IRC17:32
*** openstackgerrit has joined #openstack-barbican17:32
*** ccneill has quit IRC17:34
*** pwp has joined #openstack-barbican17:36
*** pwp has quit IRC17:38
*** pwp has joined #openstack-barbican17:38
*** mp1 has quit IRC17:43
*** edtubill has joined #openstack-barbican17:55
*** ccneill__ has joined #openstack-barbican17:57
*** ccneill_ has quit IRC18:00
*** pdesai has joined #openstack-barbican18:01
*** mp1 has joined #openstack-barbican18:04
*** silos has quit IRC18:10
*** su_zhang has joined #openstack-barbican18:20
*** dimtruck is now known as zz_dimtruck18:21
*** su_zhang has quit IRC18:22
*** su_zhang has joined #openstack-barbican18:22
*** kfarr has joined #openstack-barbican18:28
*** ccneill__ has quit IRC18:34
*** gyee has joined #openstack-barbican18:37
*** zz_dimtruck is now known as dimtruck18:38
*** gariveradlt has quit IRC18:46
*** gariveradlt has joined #openstack-barbican18:49
*** silos has joined #openstack-barbican18:53
*** ccneill__ has joined #openstack-barbican18:57
*** whydidyoustealmy has joined #openstack-barbican19:03
*** ccneill__ is now known as ccneill19:04
*** lvh_ has joined #openstack-barbican19:07
*** jkf_ has joined #openstack-barbican19:08
*** lbragstad_ has joined #openstack-barbican19:11
*** barra204 has quit IRC19:12
*** arunkant has quit IRC19:12
*** lvh has quit IRC19:12
*** jkf has quit IRC19:12
*** lbragstad has quit IRC19:12
*** lbragstad_ is now known as lbragstad19:12
*** jkf_ is now known as jkf19:12
*** silos has quit IRC19:12
*** arunkant has joined #openstack-barbican19:16
*** gariveradlt has quit IRC19:26
*** mp1 has quit IRC19:28
*** pdesai has quit IRC19:30
*** su_zhang has quit IRC19:32
*** su_zhang has joined #openstack-barbican19:32
*** silos has joined #openstack-barbican19:32
silosrellerreller: ping19:33
rellerrellersilos pong19:33
silosrellerreller: are you familiar with the glance image signing feature? Saw some apl e-mails floating around and wanted to poke your brain about something19:34
*** su_zhang has quit IRC19:34
silos*apl email usernames19:34
*** su_zhang has joined #openstack-barbican19:34
rellerrellersilos what about it?19:34
rellerrellersilos I am familiar with it.19:34
silosrellerreller. I'm confused why they create a pair of asymmetric keys. I never see the public key being used for verification. I think I might be missing something but do you know what the public key is actually used for?19:35
rellerrellersilos the image is signed with the private key. The public key/certificate is then used in two places.19:36
rellerrellersilos Glance will use the public key to verify the integrity of the image as it is uploaded to Glance.19:37
rellerrellersilos if the signature check fails then the image is not uploaded.19:37
rellerrellersilos Nova also uses the public key. When Nova is given a signed image it must first validate the signature. If the signature check fails then the image is not launched.19:38
kfarrsilos, check out this spec (problem description section) https://specs.openstack.org/openstack/glance-specs/specs/liberty/image-signing-and-verification-support.html#problem-description19:38
rellerrellersilos does that make sense?19:38
silosrellerreller: yea that makes sense. I just don't see in the tutorial here, https://etherpad.openstack.org/p/liberty-glance-image-signing-instructions, where the public key actually gets used. It seems like the private key is doing everything.19:41
siloskfarr: thanks! I've been reading through it.19:41
rellerrellersilos not sure where that came from.19:42
*** su_zhang has quit IRC19:43
*** su_zhang has joined #openstack-barbican19:43
kfarrrellerreller Those instructions came from Brianna from when she went to the liberty glance mid-cycle19:43
silosrellerreller: Okay. I'll just read through the spec more that kfarr sent. I know the pubkey gets attached to the cert on creation so it just seemed weird in that tutorial they created a public key before creating the cert. I was wondering what they were doing with it but I think it's just a step that's messing with my brain.19:44
siloskfarr, rellerreller: thanks for the help!19:46
*** su_zhang has quit IRC19:46
*** su_zhang has joined #openstack-barbican19:47
*** su_zhang has quit IRC19:53
*** whydidyoustealmy is now known as barra20419:53
*** su_zhang has joined #openstack-barbican19:53
diazjfrellerreller, kfarr, I'll setup the hangout in 5 -10 mins19:57
rellerrellerdiazjf I'm trying to wrap up things here.19:57
*** gyee has quit IRC19:57
diazjfrellerreller, sure just ping me when you get a chance19:57
diazjfedtubil, silos, jhfeng, I'd like you guys to join in as well19:58
diazjfedtubill ^19:58
edtubilldiazjf: sure19:59
rellerrellerdiazjf I'm going to sign off and be back in 2 minutes.20:00
*** rellerreller has quit IRC20:00
jhfengdiazjf: is this for BYOK ?20:00
diazjfjhfeng yup20:01
diazjfI'll post the link in 5 mins20:01
*** rellerreller has joined #openstack-barbican20:03
rellerrellerdiazjf I'm ready20:03
diazjfrellerreller perfect setting up now20:03
kfarrbrb20:05
diazjfspec: https://review.openstack.org/#/c/271517/ hangout: https://hangouts.google.com/hangouts/_/fiu.edu/byok20:05
diazjfrellerreller, kfarr, silos, edtubill, jhfeng20:05
*** su_zhang has quit IRC20:06
*** silos has quit IRC20:11
*** pwp has quit IRC20:12
*** pwp has joined #openstack-barbican20:13
redrobotarunkant ping20:16
arunkantredrobot: pong20:16
redrobotarunkant hi!  do you have a second to talk about my bug ?20:16
arunkantredorobot: Yes.20:17
redrobotarunkant ok, so I wanted to explain the bug I'm seeing in my deployment20:17
redrobotarunkant I have a pretty straightforward setup20:18
redrobotarunkant with a single Load Balancer, and a few API nodes behind it20:18
redrobotarunkant like so: http://i.imgur.com/UKycd1D.png20:18
redrobotarunkant in this case, the conf file in both API nodes is set identically such that the conf files both have the LB DNS entry as the host_href20:19
redrobotarunkant this makes the URLs all use the LB hostname, and all is well20:20
redrobotarunkant the bug I'm seeing is that when I go to get the Version, the URL I get back depends on whether the LB sent the request to api-n01 or api-n0220:20
redrobotso the Versions is returning api-n01.example.com sometimes, and other times it returns api-n0.example.com20:20
redrobot* n0220:21
redrobotobviously this is wrong, for this scenario regardless of host, the URL should always point to the LB20:21
redrobotarunkant does that make sense?20:21
*** silos has joined #openstack-barbican20:21
arunkantredrobot: So n01 or n02 are the servers which are receiving barbican requests..20:22
redrobotarunkant correct20:22
arunkantredrobot: yes it makes sense20:22
redrobotarunkant and because my LB changes the destination from the LB url to the node URL to forward it, I end up getting the wrong version back20:22
redrobot* wrong url back20:22
redrobotarunkant with the patch I proposed, the Version will also use the configuration url so regardless of n01 or n02, I will always get the right url back.  does that make sense?20:23
redrobotarunkant now, if I understand your use case, you need to be able to serve from 2 different URLS20:24
redrobotarunkant but for some reason, this needs to be the same deployment20:24
redrobotarunkant does that sound like the use case you're thinking of?20:24
diazjfjhfeng, can you join20:25
arunkantredrobot: version logic derives hostname from wsgi request..so why hostname will have n01 or n02 as that's not the client..its the server20:25
redrobotarunkant it's because haproxy is modifying the request.20:25
redrobotarunkant haproxy changes the destination to the hostname to forward the request20:26
redrobotarunkant ... well actually it's repose, but that's beside the point20:26
redrobotarunkant the point is that every single url in barbican is using the CONF value except for the version resource20:26
arunkantOkay...so its the similar to setup we have.. We set X-Forwarded-For header which then reflects correctly in hostname20:26
redrobotarunkant correct, so we're running barbican in a container, and in our setup the Version always returns the name of the container... and since it's not configurable, I have no way to fix it without the patch I submitted20:27
*** mp1 has joined #openstack-barbican20:28
*** mp1 has quit IRC20:28
redrobotarunkant now, as I understand your use case, you want to guarantee that users accessing your single deployment through the external endpoint should see the external endpoint in the URIs20:28
*** mp1 has joined #openstack-barbican20:28
redrobotarunkant and users accessing the same exact deployment through the different internal endpoint should see that same internal endpoint in the URIs, correct?20:28
arunkantredrobot: Yes..that's the concern. I was going to submit a bug and possibly fix to change href conctruction to use same logic as version controller.20:29
redrobotarunkant so, my argument is that you don't need to change barbican at all to achieve that20:29
redrobotarunkant you can achieve that today20:29
redrobotarunkant with the code as it is now, no need to change anything... except land my bugfix so that you don't see the same incorrect URL20:30
redrobotarunkant this is what your deployment should look like: http://i.imgur.com/D1uw80O.png20:30
arunkantredrobot: Okay..Can you please help me in understanding...how this change will address the issue of accessing from public vs internal client20:31
redrobotarunkant so, since you're providing two urls, I assume the client knows which endpoint to use?20:31
arunkantredrobot: client just configure/provide interface in barbicanclient and its driven via keystone version discovery using service catalog20:32
redrobotarunkant sure, so the client picks "internalUrl" or whatever20:33
arunkantright20:33
redrobotarunkant and that resolves to either external.example.com or internal.example.com in my drawing20:33
redrobotyou need to have two load balancers20:33
redrobotlistening on each of those urls20:33
redrobotand two sets of API nodes20:33
arunkantyes..two LB and same barbican API instances20:34
redrobotarunkant no, you need two different sets of API instances20:34
redrobotarunkant that is how you get the current barbican to work for you20:34
redrobotarunkant you need one set that is used exclusively by one LB, and another set that is used by the other LB20:34
arunkantredrobot: Why you need to have different instances...its the difference of which network they are coming from..20:35
redrobotin one set of API nodes, you set host_href=external.example.com in the other set of API nodes you set host_href=internal.example.com20:35
redrobotarunkant this way you can always guarantee that the API node will use the correct LB address20:36
arunkantredrobot: Are we saying any service who needs to have public and internal endpoints..they will run two of everything just to allow how client access it.20:37
redrobotarunkant yes, if you need to have two distinct endpoints, you need to deploy at least 2 api nodes20:37
redrobotarunkant note that you can still share the DB20:37
redrobotarunkant and you can still share the HSM20:38
redrobotarunkant and the workers can listen on the same queue20:38
redrobotarunkant and regardless of what endpoint the client uses, they will be able to access the same object if they provide the same UUID20:38
*** pwp has quit IRC20:38
*** pwp has joined #openstack-barbican20:38
arunkantredrobot: I am not sure that is the intention of having different interfaces in service catalog. No actual deployment will run two sets of nova/glance etc.20:39
redrobotarunkant why don't you like this solution?  nodes are cheap, that's how you scale barbican20:39
redrobotarunkant our nova deployment runs hundreds of nova nodes20:39
redrobotarunkant if you want to run a single VM with barbican you're going to have some serious performance issues20:40
arunkantredrobot: Its not about how many instances of a service...its about running different configuration, maintaining in different control plane. Its lot of baggage comes with that in terms of maintenance, upgrades etc..20:42
*** pwp has quit IRC20:44
*** pwp has joined #openstack-barbican20:44
arunkantredrobot: the problem you mentioned earlier, we have solved it by setting X-Forwarded-For header which is a standard way when proxies are involved.20:45
*** pwp has quit IRC20:45
*** pwp has joined #openstack-barbican20:46
*** edtubill has quit IRC20:47
*** pwp has quit IRC20:49
*** rellerreller has quit IRC21:00
redrobotarunkant I disagree21:07
redrobotarunkant (sorry had a meeting)21:07
redrobotarunkant I think all HREFs should behave the same way throughout the API21:07
redrobotarunkant and I insist that my fix for the Version HREF is the correct one at this time.21:07
arunkantredrobot: I think..better solution is to have href side fixed so barbican can work across many deployments which follow openstack standard endpoint interface types.21:09
*** spotz_zzz is now known as spotz21:11
redrobotarunkant you are entitled to your own opinion, however if you want to change the behavior, we should do so via a blueprint.  I think that a configurable URL is a better approach for many types of deployments21:11
arunkantredrobot: May be its something we can discuss in team meeting to get broader inputs from community21:11
redrobotarunkant the issue I'm seeing would not be fixed by X-Forwarded-For headers, since I'm using docker networking, and not just a LB in front of my APIs21:11
*** pwp has joined #openstack-barbican21:11
redrobotarunkant I think that this bugfix and your proposed change are two different conversations21:13
redrobotarunkant we can talk about your proposed change some more, but I think that the inconsistent state we're in right now is bad, and we should land my patch to fix it21:14
arunkantredrobot: I am just asking to have broader inputs ..just to see if deployments are there which have more than one endpoints (like ours).21:15
redrobotarunkant and I don't want to stop you from doing that.... I just don't think that this fix should be blocked because of that conversation21:15
*** pwp has quit IRC21:16
arunkantredrobot: From bug fix perspective, for us this will introduce bug on our side but that's never a concern.21:17
*** pwp has joined #openstack-barbican21:18
redrobotarunkant you have more problems than just this Version URL, since ALL OTHER URIs don't behave this way.21:19
redrobotarunkant and like I just explained, you can fix your deployment with a little devops using the deployment diagram I linked earlier21:19
*** su_zhang has joined #openstack-barbican21:20
*** gyee has joined #openstack-barbican21:22
arunkantredrobot: Running a parallel cloud just to differentiate public and internal clients..that not going to fly even if I want to ask them switch.21:22
*** gyee has quit IRC21:22
*** Guest40848 has joined #openstack-barbican21:23
arunkantredrobot: Anyhow, I raised my concern and thoughts about the issue. I believe that you will do the right thing.21:24
redrobotarunkant but it's not a parallel cloud.  It's just a different set of api nodes behind the second load balancer that you already have....  honestly I don't know how to convince your ops team that the deployment diagram I provided is a good solutinon :(21:24
*** pwp has quit IRC21:24
*** su_zhang has quit IRC21:25
redrobotarunkant are you striving to provide zero-downtime API access?  because your ops team is going to have to juggle different sets of api nodes to achieve that as well...21:25
*** pwp has joined #openstack-barbican21:26
arunkantredrobot: I just now looked how keystone handle that.. https://github.com/openstack/keystone/blob/master/keystone/common/wsgi.py#L37121:26
arunkantredrobot: So keystone provides the option to use which interface or it derives from wsgi request (similar to version logic)21:27
redrobotarunkant you're welcome to propose a BP for that change, but I insist that it's a separate concern from my bugfix that will make all urls consistent.21:28
*** su_zhang has joined #openstack-barbican21:32
*** pwp has quit IRC21:32
*** pwp has joined #openstack-barbican21:35
*** spotz is now known as spotz_zzz21:36
*** gyee has joined #openstack-barbican21:37
arunkantredrobot: Is it possible to discuss this in weekly team meeting or that's not an option here?21:38
*** pwp has quit IRC21:38
*** pdesai has joined #openstack-barbican21:42
*** spotz_zzz is now known as spotz21:42
*** silos has quit IRC21:43
*** panatl has joined #openstack-barbican21:44
jkfAny cores around that can look at and +2/+W my padding bug fix for the new pkcs11 code? https://review.openstack.org/#/c/27057221:46
jkfBeen sitting for nearly 2 weeks now and I think its ready to come home. :)21:47
*** silos has joined #openstack-barbican21:49
*** pwp has joined #openstack-barbican21:55
*** pwp has quit IRC21:57
*** pwp has joined #openstack-barbican21:58
*** fnaval_ has joined #openstack-barbican22:05
*** fnaval has quit IRC22:07
*** pwp has quit IRC22:12
openstackgerritElvin Tubillara proposed openstack/barbican: Simple soft deletion clean up for barbican-db-manage  https://review.openstack.org/26990322:14
*** edtubill has joined #openstack-barbican22:14
*** kebray has quit IRC22:34
arunkantHi..can anyone workflow this CR, it has 3 +2.. https://review.openstack.org/#/c/263358/22:35
*** jamielennox|away is now known as jamielennox22:41
*** silos has left #openstack-barbican22:43
*** edtubill has quit IRC22:47
*** fnaval_ is now known as fnaval22:50
*** dimtruck is now known as zz_dimtruck22:56
*** jamielennox is now known as jamielennox|away23:05
*** su_zhang has quit IRC23:11
*** su_zhang has joined #openstack-barbican23:12
*** mp1 has quit IRC23:13
*** kfarr has quit IRC23:16
*** jhfeng has quit IRC23:27
*** su_zhang has quit IRC23:51
*** su_zhang has joined #openstack-barbican23:51
« Monday, 2016-02-01 Index Wednesday, 2016-02-03 »

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!