| *** Qiming has quit IRC | 00:28 | |
| *** openstackgerrit has quit IRC | 00:46 | |
| *** openstackgerrit has joined #senlin | 00:47 | |
| *** Qiming has joined #senlin | 01:41 | |
| *** mathspanda has joined #senlin | 01:43 | |
| xuhaiwei | Qiming, morning | 02:03 |
|---|---|---|
| Qiming | hi, morning | 02:03 |
| xuhaiwei | I reported this bug, can you have a look of it? https://bugs.launchpad.net/senlin/+bug/1478753 | 02:03 |
| openstack | Launchpad bug 1478753 in senlin "Trust can't be created" [Undecided,New] | 02:03 |
| Qiming | it is about your openstacksdk version | 02:04 |
| Qiming | you will use the master instead of a public release, say 0.5.0 or 0.5.1 | 02:04 |
| xuhaiwei | I am using 0.5.1 now | 02:05 |
| Qiming | instead of a public release, say 0.5.0 or 0.5.1 | 02:05 |
| xuhaiwei | ok | 02:05 |
| Qiming | you will use the master head | 02:06 |
| Qiming | something changed recently to sdk | 02:06 |
| Qiming | xuhaiwei, need help on switching to master head? | 02:09 |
| xuhaiwei | yes | 02:10 |
| Qiming | 1. cd /usr/lib/python2.7/site-packages | 02:10 |
| Qiming | 2. rm -fr openstack | 02:10 |
| Qiming | 3. cd /opt/stack/ | 02:10 |
| Qiming | 4. git clone git://git.openstack.org/stackforge/python-openstacksdk sdk | 02:11 |
| Qiming | 5. cd sdk | 02:11 |
| Qiming | 6. sudo pip install -e . | 02:11 |
| xuhaiwei | sadly, nothing changed | 02:12 |
| xuhaiwei | $ pip list|grep openstacksdk openstacksdk (0.5.1.dev236, /opt/stack/python-openstacksdk) | 02:12 |
| xuhaiwei | this is the current version | 02:13 |
| Qiming | ok | 02:13 |
| xuhaiwei | my openstacksdk is installed to /usr/local/lib/python2.7/dist-packages by default | 02:13 |
| xuhaiwei | I usually reinstalled it by running 'python setup.py install ', but this doesn't work either | 02:14 |
| Qiming | when you do that install, you will be copying code into /usr/lib/... | 02:15 |
| Qiming | since we will need to find sdk errors frequently, I'd suggest you do an install under /opt/stack/ | 02:15 |
| Qiming | "sudo pip install -e ." will only create a "openstacksdk.egg-link" file under /usr/lib/python2.7 | 02:16 |
| *** Zhenqi has joined #senlin | 02:16 | |
| Qiming | that way, you will be sure you can change sdk code and make it work if there are bugs found | 02:17 |
| *** Tennyson has joined #senlin | 02:17 | |
| xuhaiwei | now in /usr/local/lib/python2.7/disk-packages, I have openstacksdk-0.5.1.dev236-py2.7.egg-info/ | 02:18 |
| Qiming | that one should be deleted as well | 02:18 |
| Qiming | that file was created because you did "python setup.py install" | 02:19 |
| xuhaiwei | I have deleted all the related files in under /usr/local/lib/python2.7/dist-packages, and reinstalled openstacksdk as you said, but it still has the same error | 02:23 |
| Qiming | checking | 02:24 |
| Qiming | did you do a reinstall of senlin? | 02:25 |
| Qiming | just found a problem | 02:26 |
| xuhaiwei | Yesterday, in order to test this bug report https://launchpad.net/bugs/1478423, Yanyan asked me to delete the credential from the db | 02:26 |
| openstack | Launchpad bug 1478423 in senlin " HttpException: 403 Client Error: Forbidden, At least one role should be specified" [High,Triaged] | 02:26 |
| Qiming | in 0.5.1 or so, the distribution name was changed from python-openstacksdk to just 'openstacksdk' | 02:27 |
| xuhaiwei | After i deleted the credentials, I got this problem | 02:27 |
| Qiming | okay, I'm trying | 02:28 |
| Qiming | okay, I'm seeing some problems, not exactly the same one as yours | 02:29 |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: Fix trust listing https://review.openstack.org/206282 | 02:43 |
| Qiming | xuhaiwei, just fixed a problem in trust listing | 02:43 |
| xuhaiwei | can resolve my problem? | 02:46 |
| xuhaiwei | Qiming, the problem is solved | 02:52 |
| Qiming | yours is solved? | 02:53 |
| xuhaiwei | but there are two versions openstacksdk | 02:53 |
| Qiming | yep | 02:53 |
| xuhaiwei | yes, the problem I reported is resolved | 02:53 |
| Qiming | okay, we only use the latest version | 02:53 |
| Qiming | considering that they will do a release soon | 02:54 |
| xuhaiwei | how does senlin know which openstacksdk to use? | 02:54 |
| xuhaiwei | I think senlin is using the one installed under /usr/local/lib/python2.7/dist-packages | 02:56 |
| Qiming | with that openstacksdk 1.0 is release, we can propose that to the global-requirements | 02:58 |
| Qiming | now the requirements.txt file is not good | 02:58 |
| Qiming | it says "python-openstacksdk>=0.4.1" | 02:58 |
| Qiming | if there is no such a package installed, senlin will ask for one | 02:58 |
| xuhaiwei | I am a little confused about this, in that case, senlin can't use the latest version I think | 02:59 |
| Qiming | we should replace that line with openstacksdk>=0.5.0 | 02:59 |
| Qiming | but that will introduce a conflict with the global requirements | 02:59 |
| Qiming | get the points? | 02:59 |
| Qiming | because you did not follow the steps I show you | 02:59 |
| xuhaiwei | your explanation above makes sense to me | 03:00 |
| Qiming | this is an uneasy period for both projects | 03:01 |
| xuhaiwei | yes, you mean I should delete the old version? | 03:01 |
| Qiming | we hope sdk will do a release soon | 03:01 |
| Qiming | right | 03:01 |
| Qiming | delete the old version | 03:01 |
| xuhaiwei | ok | 03:01 |
| Qiming | or else you will be mixing code versions | 03:03 |
| xuhaiwei | yes | 03:03 |
| xuhaiwei | but everytime we run senlin/pip install -e ., we will get the old version again | 03:03 |
| Qiming | <Qiming> now the requirements.txt file is not good | 03:04 |
| Qiming | <Qiming> it says "python-openstacksdk>=0.4.1" | 03:04 |
| Qiming | <Qiming> if there is no such a package installed, senlin will ask for one | 03:04 |
| Qiming | <Qiming> we should replace that line with openstacksdk>=0.5.0 | 03:04 |
| xuhaiwei | locally? | 03:05 |
| Qiming | yes | 03:06 |
| Qiming | you won't be able to commit it | 03:06 |
| Qiming | because the gate will refuse it | 03:06 |
| xuhaiwei | ok | 03:06 |
| *** mathspanda has quit IRC | 03:41 | |
| *** mathspanda has joined #senlin | 04:18 | |
| xuhaiwei | Qiming, I'd like to implement the parse_sdk_exception method, you are not working on it, right? | 04:57 |
| Qiming | no, I'm not, xuhaiwei | 05:10 |
| xuhaiwei | ok, should we add it in senlin/common/utils.py, or somewhere else? | 05:11 |
| xuhaiwei | senlin/common/excutils? | 05:11 |
| Qiming | senlin/drivers/openstack/sdk.py | 05:12 |
| Qiming | there is a partial implementation there | 05:12 |
| xuhaiwei | ok | 05:13 |
| *** mathspanda has quit IRC | 05:14 | |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: Fix some time related function calls https://review.openstack.org/206318 | 06:03 |
| lixinhui | Qiming, is any brenchmark you know used by heat or someone else for scale out/in scenarios? | 06:19 |
| Qiming | lixinhui, no | 06:19 |
| Qiming | we test it by logging into each individual servers and burn the CPU | 06:20 |
| Qiming | I've just noticed that someone has proposed a different approach here: https://review.openstack.org/#/c/205939/1/hot/mistral/autoscaling_using_mistral.yaml | 06:21 |
| Qiming | it is about generating a gzip archive from urandom stream | 06:22 |
| lixinhui | :) | 06:24 |
| xuhaiwei | it's using mistral to do auto-scaling? | 06:24 |
| lixinhui | Qiming, what you use to burn cpu? | 06:24 |
| Qiming | xuhaiwei, it is using mistral to do .... cpu burning | 06:25 |
| Qiming | that is ... funny | 06:25 |
| xuhaiwei | instead of ceilometer? | 06:26 |
| Qiming | ceilometer doesn't do cpu burning | 06:26 |
| xuhaiwei | cpu burning means consuming the cpu? | 06:30 |
| Qiming | yes | 06:30 |
| Qiming | push CPU utilization to a high value so it will be deteced by ceilometer and thus notified to Heat | 06:30 |
| xuhaiwei | just for test? | 06:31 |
| Qiming | yes | 06:31 |
| Qiming | I was expecting someone to show the whole process modeled as a mistral workflow | 06:31 |
| Qiming | the sample template above is disappointing | 06:31 |
| Zhenqi | node_create has met a problem | 06:51 |
| Zhenqi | in nova/server.py | 06:52 |
| Zhenqi | aboout image.id | 06:52 |
| Zhenqi | and our openstacksdk is the latest | 06:53 |
| Qiming | Zhenqi, this info is not helpful for a diagnosis | 06:53 |
| Qiming | please be more specific | 06:54 |
| Zhenqi | wrong information " AttributeError: 'NoneType' object has no attribute 'id'" | 06:54 |
| Qiming | any info about the line number? | 06:55 |
| Zhenqi | line 220 | 06:55 |
| Zhenqi | File "/opt/stack/senlin/senlin/profiles/os/nova/server.py", line 220, in do_create | 06:56 |
| openstackgerrit | xu-haiwei proposed stackforge/senlin: Refactor random heat stack name https://review.openstack.org/206332 | 06:57 |
| xuhaiwei | Zhenqi, paste all the information here http://paste.openstack.org/ | 06:57 |
| Zhenqi | OK | 06:58 |
| xuhaiwei | gate seems very slow today | 06:59 |
| Qiming | -openstackstatus/#openstack-dev- NOTICE: zuul is stuck and about to undergo an emergency restart, please be patient as job results may take a long time | 06:59 |
| Qiming | Zhenqi, did you paste the information? | 07:04 |
| Qiming | you will need to tell us the url where the info is pasted | 07:04 |
| Zhenqi | http://paste.openstack.org/show/406010/ | 07:06 |
| Qiming | Zhenqi, you may have specified an image you don't have | 07:08 |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: Fix image finding call https://review.openstack.org/206336 | 07:08 |
| Qiming | there are things Senlin can improve, see the patch above ^ | 07:08 |
| Qiming | however, the root cause of the error you encountered is about image not found | 07:09 |
| xuhaiwei | the exception handling is not good here :) | 07:10 |
| Qiming | ya, need to improve it | 07:10 |
| Qiming | if the image is not found, we were expecting SDK to raise a 404 | 07:10 |
| xuhaiwei | yes | 07:10 |
| Qiming | SDK, however, is not doing that, if we don't specify ignore_missing=False | 07:11 |
| xuhaiwei | sdk just returns None? | 07:11 |
| Qiming | seems so | 07:11 |
| Qiming | and the 'ignore_missing' parameter was just added recently | 07:11 |
| xuhaiwei | for us, all the ignore_missing parameters are desired to be set False | 07:12 |
| Zhenqi | In fact, I found the image in nova image-list | 07:14 |
| xuhaiwei | can you show your profile spec Zhenqi? | 07:15 |
| Zhenqi | sorry | 07:18 |
| Zhenqi | the profile example use an old version cirros-0.3.2-x86_64-uec | 07:18 |
| *** Zhenqi is now known as Xiangyang | 07:25 | |
| lixinhui | Qiming, are you using cpuburn in your own tests? | 07:31 |
| Qiming | no | 07:31 |
| Qiming | Haven't touch that for some time | 07:31 |
| Qiming | a simple one would be an infinite loop doing some useless computations | 07:32 |
| lixinhui | that must be elegantly written for multi-core servers | 07:33 |
| lixinhui | as stress test | 07:33 |
| Qiming | e.g. | 07:36 |
| Qiming | for cpu in 1 2 3 4; do | 07:36 |
| Qiming | ( while true; do true; done ) & | 07:36 |
| Qiming | done | 07:36 |
| Qiming | a single line can be: | 07:37 |
| Qiming | cat /dev/urandom | gzip > /dev/null | 07:37 |
| lixinhui | http://www.hecticgeek.com/2012/03/cpuburn-cpu-stress-test-ubuntu-linux/ | 07:40 |
| Qiming | well, if you want some native applications for stress testing, you can find many of them | 07:45 |
| openstackgerrit | Merged stackforge/senlin: Fix test case of environment https://review.openstack.org/205900 | 07:50 |
| openstackgerrit | Merged stackforge/senlin: Transfer roles of context as inputs to create trust https://review.openstack.org/205874 | 07:50 |
| *** mathspanda has joined #senlin | 08:12 | |
| lixinhui | Qiming, what scenarios are targeted by heat anti-affinity | 08:15 |
| Qiming | HA | 08:18 |
| lixinhui | okay,haha | 08:21 |
| lixinhui | do they have requirement to affinity? | 08:22 |
| lixinhui | or already have scenarios | 08:22 |
| Qiming | who are they> | 08:26 |
| lixinhui | HEAT | 08:28 |
| lixinhui | or HEAT users | 08:28 |
| Qiming | certainly | 08:28 |
| lixinhui | what scenarios are targeted by heat affinity? | 08:29 |
| Qiming | heat doesn't support this yet | 08:30 |
| lixinhui | design reason? | 08:31 |
| Qiming | not in its scope | 08:35 |
| lixinhui | :) | 08:38 |
| lixinhui | it seems that sahara used anti-affinity of HEAT | 08:44 |
| *** Tennyson has quit IRC | 09:01 | |
| *** Xiangyang has quit IRC | 09:11 | |
| Qiming | Heat provides no anti-affinity | 09:13 |
| *** mathspanda has quit IRC | 09:18 | |
| openstackgerrit | Merged stackforge/senlin: Refactor random heat stack name https://review.openstack.org/206332 | 09:53 |
| *** Qiming has quit IRC | 10:15 | |
| openstackgerrit | Merged stackforge/senlin: Fix image finding call https://review.openstack.org/206336 | 11:01 |
| *** Qiming has joined #senlin | 11:07 | |
| lixinhui | Qiming, there? | 11:09 |
| Qiming | yes | 11:12 |
| lixinhui | Qiming, sorry for the delay, on call just now. | 11:54 |
| lixinhui | I saw heat code under engin/nova, and sahara template includes the realted properties | 11:56 |
| lixinhui | is there any discussion in heat community on affinity and antiaffinity | 11:56 |
| lixinhui | how for these people to decide not in heat's scope | 11:57 |
| Qiming | if you are talking about scheduler_hints | 12:05 |
| Qiming | they are just properties exposed by nova | 12:05 |
| Qiming | heat's job is to expose those properties faithfuly to template writers | 12:06 |
| Qiming | heat doesn't care about what the properties mean | 12:06 |
| lixinhui | agree, please share more discussion in heat about affintiy/antiaffinity | 12:06 |
| Qiming | heat doesn't want to get involved in any scheduling decisions | 12:06 |
| Qiming | there is no such discussion in heat | 12:07 |
| Qiming | you have to work with the team for a long time to understand their philosophy | 12:07 |
| lixinhui | wanna follow this: lixinhui> Qiming, what scenarios are targeted by heat anti-affinity | 12:08 |
| lixinhui | <Qiming> HA | 12:08 |
| Qiming | yes | 12:08 |
| Qiming | please do | 12:09 |
| lixinhui | maybe I just misunderstand what you mean by HA, did you mean haha that time? | 12:09 |
| Qiming | no, high-availability | 12:10 |
| Qiming | anti-affinity is just a tiny component of a HA solution | 12:12 |
| Qiming | a full HA solution involves: 1) failure detection 2) failure notification 3) recovery | 12:13 |
| Qiming | there are many factors related to the failure detection component, for example | 12:14 |
| Qiming | host failure, network failure, storage failure, VM failure and application failure | 12:14 |
| Qiming | each of which needs a solution for reliable, deterministic detection | 12:15 |
| Qiming | if you look into recovery, there are many factors there as well | 12:15 |
| Qiming | you will need to make sure the failed nodes are fenced before taking recovery measures | 12:16 |
| Qiming | you will need to choose from many alternatives (maybe in sequence) for recovery | 12:16 |
| Qiming | for a VM, the recovery action could be: reboot, rebuild, evacuate, migrate, recreate | 12:17 |
| lixinhui | Thanks for sharing, Qiming | 12:20 |
| Qiming | I have spent about one year on high-availability | 12:20 |
| lixinhui | how anti-affinity will be involved in the loop? | 12:20 |
| Qiming | a lot to share, ;) | 12:20 |
| lixinhui | I believe :) | 12:21 |
| Qiming | anti-affinity is just a preparation step for the loop | 12:21 |
| lixinhui | I see, you mean diffreciate the group need HA and not | 12:21 |
| Qiming | no | 12:22 |
| Qiming | a group would be better placed using a policy like anti-affinity | 12:22 |
| Qiming | so that the service continuity can be guaranteed to a certain degree | 12:22 |
| *** jruano has joined #senlin | 12:23 | |
| lixinhui | okay, got it | 12:24 |
| Qiming | I have presentations on the Atlanta summit and demos on the Paris summit | 12:26 |
| lixinhui | do you have charts of the two talk? and vedio if record available | 12:27 |
| Qiming | cannot access the video | 12:28 |
| Qiming | but I do have charts for both | 12:28 |
| lixinhui | send them to me, xiaoming, you are a trasure | 12:30 |
| lixinhui | trasure | 12:30 |
| lixinhui | treasure | 12:30 |
| lixinhui | something wrong with 'e' on my keyboard, I must tap it heavily to input 'e' | 12:32 |
| Qiming | use your toes | 12:32 |
| Qiming | remind me later, busy now | 12:32 |
| lixinhui | okay, I will try ... | 12:34 |
| *** xuhaiwei_ has joined #senlin | 12:39 | |
| *** Yanyan has joined #senlin | 12:40 | |
| *** lkarm has joined #senlin | 12:45 | |
| *** Yanyanhu has joined #senlin | 12:46 | |
| *** Yanyan has quit IRC | 12:48 | |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: Sample triggers using ceilometer alarm https://review.openstack.org/206526 | 12:49 |
| *** Yanyan has joined #senlin | 12:50 | |
| *** Yanyanhu has quit IRC | 12:51 | |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: WIP: Trigger for ceilometer alarms https://review.openstack.org/206527 | 12:52 |
| Yanyan | meeting time? | 13:00 |
| lixinhui | yes, I am in | 13:02 |
| openstackgerrit | Yanyan Hu proposed stackforge/senlin: Revise some TODO items https://review.openstack.org/206542 | 13:42 |
| Qiming | hoho, I didn't count down to 1 | 13:56 |
| Yanyan | finally finish the re-installation of my environment | 13:56 |
| Yanyan | ... | 13:56 |
| xuhaiwei_ | it is too hot here | 13:56 |
| Yanyan | sufferred so much from pbr version conflict... | 13:57 |
| xuhaiwei_ | 1.3.0 should be ok i think | 13:57 |
| Yanyan | have to do too much manual interference | 13:59 |
| Yanyan | beijing is also very hot | 13:59 |
| Yanyan | just found serveral packages are still using pbr with lower version | 13:59 |
| Yanyan | have to manually handle them | 13:59 |
| Yanyan | and then the cliff package version can't be satisfied... | 13:59 |
| xuhaiwei_ | anyway you solved it | 14:00 |
| Yanyan | yes :) | 14:00 |
| xuhaiwei_ | i will quit ,see you guys tomorrow | 14:00 |
| xuhaiwei_ | good night | 14:00 |
| Yanyan | good night | 14:00 |
| Yanyan | hi, Qiming, about this patch https://review.openstack.org/170355 | 14:01 |
| Yanyan | need we merge it? I thought it was a test code but not sure whether we need it. | 14:02 |
| Qiming | em, totally forget what it is about :( | 14:02 |
| Qiming | ah, I see | 14:02 |
| Qiming | it is not very useful I think | 14:03 |
| Qiming | it was an attempt to remove request context from invocation path | 14:03 |
| Yanyan | yes | 14:03 |
| Qiming | feel free to abandon it | 14:04 |
| Yanyan | ok, or we just leave it there for possible feature usage | 14:04 |
| *** xuhaiwei_ has quit IRC | 14:04 | |
| Yanyan | I will also leave, talk to you guys later | 14:05 |
| lixinhui | Qiming, send me your decks for the two talk | 14:08 |
| *** Yanyan has quit IRC | 14:11 | |
| Qiming | ok | 14:47 |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: WIP: Trigger for ceilometer alarms https://review.openstack.org/206527 | 14:52 |
| openstackgerrit | Merged stackforge/senlin: Revise some TODO items https://review.openstack.org/206542 | 14:54 |
| openstackgerrit | Qiming Teng proposed stackforge/senlin: Sample triggers using ceilometer alarm https://review.openstack.org/206526 | 15:04 |
| *** Qiming has quit IRC | 15:32 | |
| *** lkarm has quit IRC | 18:33 | |
| *** lkarm has joined #senlin | 18:34 | |
| *** jruano has quit IRC | 21:07 | |
| *** lkarm has quit IRC | 21:43 | |
| *** jruano has joined #senlin | 21:51 | |
| *** jruano has quit IRC | 22:05 | |
| *** jruano has joined #senlin | 22:13 | |
| *** jruano has quit IRC | 22:46 | |
| *** jruano has joined #senlin | 22:59 | |
| *** jruano has quit IRC | 23:09 | |
| *** Qiming has joined #senlin | 23:31 | |
| *** Qiming has quit IRC | 23:36 | |
| *** Qiming has joined #senlin | 23:36 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!