| *** byeager has quit IRC | 00:00 | |
| *** kevinc_ has joined #openstack-swift | 00:00 | |
| *** mwstorer has quit IRC | 00:03 | |
| *** matsuhashi has joined #openstack-swift | 00:20 | |
| *** matsuhashi has quit IRC | 00:32 | |
| *** matsuhashi has joined #openstack-swift | 00:33 | |
| *** Edward-Zhang has quit IRC | 00:34 | |
| *** tnewsome has quit IRC | 00:39 | |
| *** kevinc_ has quit IRC | 00:44 | |
| *** adrian_otto has quit IRC | 00:45 | |
| *** shri has quit IRC | 00:51 | |
| *** adrian_otto has joined #openstack-swift | 00:57 | |
| *** notmyname has quit IRC | 01:03 | |
| *** adrian_otto has left #openstack-swift | 01:03 | |
| *** patchbot has quit IRC | 01:06 | |
| *** patchbot has joined #openstack-swift | 01:06 | |
| *** notmyname has joined #openstack-swift | 01:06 | |
| clayg | rebase train! | 01:14 |
|---|---|---|
| notmyname | choo choo | 01:16 |
| *** saschpe has quit IRC | 01:20 | |
| *** saschpe has joined #openstack-swift | 01:21 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Keep status_changed_at up-to-date with status changes. https://review.openstack.org/86714 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend direct_client https://review.openstack.org/88144 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add reconciler probetest outline https://review.openstack.org/90934 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policies to container.backend https://review.openstack.org/90935 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update FakeLogger with better log level handling https://review.openstack.org/90932 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add write_fake_ring helper to test.unit https://review.openstack.org/88151 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add container-reconciler daemon https://review.openstack.org/85908 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Make FakeRing a use more real code https://review.openstack.org/90933 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add LRUCache to common.utils https://review.openstack.org/90938 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend interface on InternalClient https://review.openstack.org/88152 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy to container put/delete object https://review.openstack.org/90936 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Enqueue misplaced objects during container replication https://review.openstack.org/90937 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Merge container storage_policy_index values. https://review.openstack.org/86720 | 01:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Expose container info on deleted containers. https://review.openstack.org/86708 | 01:24 |
| *** Honghui has joined #openstack-swift | 01:33 | |
| *** bill_az has quit IRC | 01:34 | |
| *** nosnos has joined #openstack-swift | 01:43 | |
| *** notmyname has quit IRC | 01:54 | |
| *** notmyname has joined #openstack-swift | 01:54 | |
| portante | clayg: are we planning on merging the in-process functional test work onto the feature/ec branch? | 01:55 |
| *** Edward-Zhang has joined #openstack-swift | 01:55 | |
| *** ChanServ sets mode: +v notmyname | 01:55 | |
| portante | the non-voting in-process func tests are failing on feature/ec because tox.ini does not have a "func" env | 01:56 |
| notmyname | portante: yes, but other way around | 01:56 |
| notmyname | portante: master will be merged into feature/ec (peluse has worked on that today) and then it will pick up the right thing | 01:57 |
| portante | ah, yes, thanks, sorry | 01:57 |
| portante | shall we rename this effort to pretzel? | 01:57 |
| notmyname | :-) | 01:57 |
| portante | ;) | 01:57 |
| notmyname | that sounds like a cool idea: "A rack level magsafe adaptor for 60amp" | 01:58 |
| portante | you got that from a pretzel? | 01:59 |
| portante | ;) | 01:59 |
| portante | true genius! | 01:59 |
| notmyname | a retweet from the head of bloomberg cloud IT :-) | 02:01 |
| notmyname | kind of a cool idea though. probably impractical | 02:02 |
| peluse_ | portante: yeah, I got the merge done (notnmyname has to push it though) and verified in process func tests working. Just had to add a 2nd obj ring to support the policy tets (2nd ring file) | 02:03 |
| notmyname | peluse_: technically, any core can push it | 02:04 |
| notmyname | for example, portante ;-) | 02:04 |
| peluse_ | heh | 02:08 |
| *** klrmn has left #openstack-swift | 02:09 | |
| portante | what is the commit id? | 02:12 |
| portante | peluse_: ^^^ | 02:12 |
| peluse_ | portante: the way we've been doing it is I merge and then push to a github then provide that to notmyname and he does the rest... | 02:14 |
| portante | oh | 02:14 |
| peluse_ | yeah, permissions issues keep me from proposing a merge I guess. We tried it a while back and it didn't work | 02:15 |
| portante | hmm, not familiar with this process at all, if what you are talking about does not go through gerrit | 02:16 |
| peluse_ | portante: no, it does just not by me. I resolve the conflicts and push to github then notmyname or torgomatic does some magic and it shows up as a gerrit deal | 02:16 |
| portante | oh, you can do that. | 02:17 |
| portante | that does not take any privs, if I understand things correctly | 02:17 |
| peluse_ | portante: hey man, I just do what I'm told :) | 02:17 |
| portante | what I do is add the git hub repo as another remote | 02:17 |
| portante | git remote add ... | 02:17 |
| peluse_ | yup, did that | 02:17 |
| portante | and then just rebase to feature/ec | 02:18 |
| peluse_ | the npush to a repo I create and share that info - just sent it to you in case you want to look at it | 02:18 |
| portante | okay | 02:18 |
| peluse_ | oh, hmm. | 02:18 |
| peluse_ | well, if someone wants me to I can take a next step - just don't want to jack it up as I'm no git gerrit wiz | 02:19 |
| peluse_ | goota bolt now though... catch ya later | 02:19 |
| portante | k | 02:19 |
| *** Edward-Zhang has quit IRC | 02:33 | |
| notmyname | portante: no, it's a merge commit, and those are restricted in gerrit to the swift-core group | 02:34 |
| notmyname | portante: https://wiki.openstack.org/wiki/GerritJenkinsGithub#Merge_Commits | 02:35 |
| *** NM1 has joined #openstack-swift | 02:38 | |
| *** praveenkumar has joined #openstack-swift | 02:46 | |
| *** zhiyan_ is now known as zhiyan | 02:53 | |
| portante | okay, thanks | 02:56 |
| *** russell_h has quit IRC | 03:01 | |
| *** NM1 has quit IRC | 03:14 | |
| *** cheri has joined #openstack-swift | 03:24 | |
| *** cheri has quit IRC | 03:30 | |
| *** matsuhashi has quit IRC | 03:33 | |
| *** gyee has quit IRC | 03:34 | |
| *** haomaiwa_ has joined #openstack-swift | 03:41 | |
| *** Honghui has quit IRC | 03:42 | |
| *** nosnos has quit IRC | 03:43 | |
| *** changbl has joined #openstack-swift | 03:48 | |
| *** NM1 has joined #openstack-swift | 03:50 | |
| *** Honghui has joined #openstack-swift | 03:55 | |
| *** NM1 has quit IRC | 04:00 | |
| *** ppai has joined #openstack-swift | 04:01 | |
| *** Midnightmyth has joined #openstack-swift | 04:07 | |
| *** Honghui has quit IRC | 04:10 | |
| openstackgerrit | Zhang Hua proposed a change to openstack/swift: Add profiling middleware in Swift https://review.openstack.org/53270 | 04:18 |
| *** ams0 has quit IRC | 04:20 | |
| *** Longgeek_ has joined #openstack-swift | 04:22 | |
| *** ams0 has joined #openstack-swift | 04:28 | |
| *** tzumainn has quit IRC | 04:34 | |
| *** matsuhashi has joined #openstack-swift | 04:39 | |
| *** chandan_kumar has joined #openstack-swift | 04:50 | |
| *** nosnos has joined #openstack-swift | 04:51 | |
| *** Midnightmyth has quit IRC | 04:52 | |
| *** Honghui has joined #openstack-swift | 04:52 | |
| *** Longgee__ has joined #openstack-swift | 04:54 | |
| *** nshaikh has joined #openstack-swift | 04:54 | |
| *** Longgeek_ has quit IRC | 04:55 | |
| *** Honghui has quit IRC | 05:04 | |
| *** Honghui has joined #openstack-swift | 05:06 | |
| *** zaitcev has quit IRC | 05:16 | |
| *** bach has joined #openstack-swift | 05:18 | |
| *** ams0 has quit IRC | 05:22 | |
| *** ams0 has joined #openstack-swift | 05:23 | |
| Honghui | Anyone can recommend me hardware specification for running Swift. We do some testing on Dell R720xd+H310 RAID card in JBOD mode, it's performance is terrible. | 05:23 |
| notmyname | Honghui: where is the bottleneck? cpu, memory, disk IO, network? | 05:23 |
| Honghui | The disk IO is bottleneck. The performance looks not bad if we configure each disk as RAID0. | 05:23 |
| *** ams0 has quit IRC | 05:25 | |
| *** CrackerJackMack has quit IRC | 05:37 | |
| *** sudorandom has quit IRC | 05:39 | |
| *** bach has quit IRC | 05:48 | |
| *** swills has quit IRC | 05:55 | |
| *** bach has joined #openstack-swift | 05:57 | |
| *** swills has joined #openstack-swift | 05:59 | |
| *** swills has quit IRC | 06:00 | |
| *** swills has joined #openstack-swift | 06:00 | |
| *** PradeepChandani has joined #openstack-swift | 06:02 | |
| Honghui | notmyname: The disk IO/RAID Card is bottleneck. The performance looks not bad if we configure each disk as RAID0. | 06:03 |
| notmyname | Honghui: are you running account, container, and object servers all on the same drives? | 06:04 |
| Honghui | notmyname: no, we running account and container in another SSD disks which is configured as RAID1. We can see significant different disk throughout between JBOD mode and RAID mode from Graphite/Statsd. We don't think it's a container update performance issue. | 06:08 |
| notmyname | ah, ok. you're way ahead of me then :-) | 06:09 |
| notmyname | Honghui: have you looking into the settings on the RAID card? eg buffer sizes, write barriers turned on or off, are they green drives trying to spin down, etc | 06:12 |
| Honghui | notmyname: We has test Dell R710 + Dataon DNS-1400, it works well, but we can't found a Dataon provider in China. It seems only Dell R720+LSI 9207-8e+MD1200 is available choice for me. | 06:12 |
| *** ams0 has joined #openstack-swift | 06:15 | |
| Honghui | notmyname: We has use the H310 card and H710 card as compare. For H310, it does not distanct write through mode and write back mode and it strip size is 64K. When we testing H710 with write back mode, the performance is same as H310 RAID mode. We are testing H710 with write through mode too see whether the buffer take a role in the performance difference. | 06:19 |
| *** mlipchuk has joined #openstack-swift | 06:22 | |
| *** ams0 has quit IRC | 06:25 | |
| notmyname | Honghui: I'm not too familiar with those cards. I'm about to go to bed, but hugokuo (who's on your side of the world) can help you some, I think | 06:25 |
| *** ams0 has joined #openstack-swift | 06:26 | |
| Honghui | notmyname: thanks! :-) | 06:26 |
| *** ams0 has quit IRC | 06:27 | |
| Honghui | hi hugokuo | 06:28 |
| hugokuo | Honghui: Hi | 06:28 |
| Honghui | hugokuo: notmyname mentioned you that you can help on decide which hardware are good to running swift. | 06:32 |
| notmyname | Honghui: sorry, I think hugokuo is eating lunch actually :-) | 06:32 |
| hugokuo | notmyname: I've done my lunch. Skype with my colleague | 06:35 |
| hugokuo | Honghui: So what's your 1) use cases 2) expected performance result. | 06:35 |
| hugokuo | Honghui: 如果需要我們可以用中文, 也許可以更精確 | 06:37 |
| Honghui | hugokuo: We are store billions of files, each file is about 128KB. each day upload 1m files and download 4m files. | 06:38 |
| *** haomaiwa_ has quit IRC | 07:06 | |
| *** psharma has joined #openstack-swift | 07:10 | |
| *** foexle has joined #openstack-swift | 07:25 | |
| *** zhiyan is now known as zhiyan_ | 07:26 | |
| chmouel | Honghui: that's an impressive number :) | 07:44 |
| omame | chmouel: that's china :) | 07:59 |
| omame | Honghui: have you tested any perc 8xx series controller? if you're looking for better i/o you should probably scale up your controller. also, the kind of usage you're doing sounds very stressful for the drive heads. have you tested different hard drive models and firmwares? | 08:02 |
| *** nacim has joined #openstack-swift | 08:11 | |
| openstackgerrit | Yuan Zhou proposed a change to openstack/swift: Allow to deprecate policy https://review.openstack.org/86827 | 08:15 |
| chmouel | omame: :) | 08:18 |
| *** mmcardle has joined #openstack-swift | 08:22 | |
| *** bach has quit IRC | 08:24 | |
| Honghui | omame: Do you means configure RAID0 on each disk? | 08:25 |
| Honghui | omame: We got performance issue when and only when we configure H310 to run in JBOD mode. | 08:26 |
| omame | my questions were more: have you tried any 8xx model in single-drive raid0 mode? have you tried different hard drive models? have you tried using the latest firmware release for your hard drives? | 08:32 |
| *** jamie_h has joined #openstack-swift | 08:35 | |
| *** nacim has quit IRC | 08:54 | |
| *** acoles_away is now known as acoles | 08:58 | |
| *** nshaikh has quit IRC | 08:58 | |
| *** mmcardle has quit IRC | 09:03 | |
| *** mmcardle has joined #openstack-swift | 09:03 | |
| *** nacim has joined #openstack-swift | 09:07 | |
| *** Honghui has quit IRC | 09:11 | |
| *** sudorandom has joined #openstack-swift | 09:12 | |
| *** CrackerJackMack has joined #openstack-swift | 09:12 | |
| *** joeljwright has joined #openstack-swift | 09:16 | |
| *** joeljwright has quit IRC | 09:16 | |
| *** chandan_kumar has quit IRC | 09:17 | |
| *** praveenkumar has quit IRC | 09:17 | |
| *** praveenkumar has joined #openstack-swift | 09:23 | |
| *** chandan_kumar has joined #openstack-swift | 09:23 | |
| *** praveenkumar has quit IRC | 09:27 | |
| *** acoles is now known as acoles_away | 09:38 | |
| *** acoles_away is now known as acoles | 09:40 | |
| *** matsuhashi has quit IRC | 09:41 | |
| *** praveenkumar has joined #openstack-swift | 09:41 | |
| *** matsuhas_ has joined #openstack-swift | 09:42 | |
| *** Honghui has joined #openstack-swift | 09:44 | |
| *** mkollaro has joined #openstack-swift | 09:45 | |
| *** chandan_kumar has quit IRC | 09:46 | |
| *** chandan_kumar has joined #openstack-swift | 09:59 | |
| *** nshaikh has joined #openstack-swift | 10:00 | |
| *** haomaiwa_ has joined #openstack-swift | 10:03 | |
| *** tzumainn has joined #openstack-swift | 10:04 | |
| jd__ | chmouel: any progress on swiftclient and Python 3 btw? | 10:08 |
| *** matsuhas_ has quit IRC | 10:09 | |
| chmouel | jd__: there is only one review left that is waiting for a +2 | 10:09 |
| chmouel | jd__: https://review.openstack.org/#/c/88351/ | 10:10 |
| * jd__ stares at all core devs | 10:10 | |
| chmouel | cschwede: ^ | 10:10 |
| jd__ | "the time has come to push the button" | 10:10 |
| *** matsuhashi has joined #openstack-swift | 10:10 | |
| chmouel | i guess it need a rebase first | 10:10 |
| jd__ | click on the rebase button? | 10:11 |
| openstackgerrit | Chmouel Boudjnah proposed a change to openstack/python-swiftclient: Only encode metadata for user customed headers https://review.openstack.org/88351 | 10:11 |
| chmouel | jd__: that wasn't workng, it needed a 3-way automatic merge which gerrit does not seem to handle | 10:11 |
| jd__ | thanks chmouel | 10:14 |
| *** matsuhashi has quit IRC | 10:21 | |
| *** matsuhashi has joined #openstack-swift | 10:22 | |
| *** acoles is now known as acoles_away | 10:27 | |
| *** NM1 has joined #openstack-swift | 10:32 | |
| *** joeljwright has joined #openstack-swift | 10:34 | |
| *** acoles_away is now known as acoles | 10:34 | |
| *** NM1 has quit IRC | 10:43 | |
| *** NM1 has joined #openstack-swift | 10:48 | |
| acoles | chmouel: ping | 10:51 |
| *** NM1 has quit IRC | 10:53 | |
| *** Longgee__ has quit IRC | 11:03 | |
| *** tdasilva has joined #openstack-swift | 11:04 | |
| hugokuo | ping hugokuo ... | 11:05 |
| hugokuo | err ... | 11:05 |
| ctennis | are you talking to yourself hugokuo? :) | 11:06 |
| *** ppai has quit IRC | 11:06 | |
| *** Honghui has quit IRC | 11:07 | |
| *** Honghui has joined #openstack-swift | 11:11 | |
| *** nacim has quit IRC | 11:16 | |
| *** ppai has joined #openstack-swift | 11:23 | |
| *** joeljwright has quit IRC | 11:30 | |
| *** kevinc_ has joined #openstack-swift | 11:33 | |
| Honghui | omame: We has compare H310 and H710. H310 in JBOD mode/RAID0 mode. H710 in RAID0 mode(Write Back/Write Through). The time: H310(JBOD): 1500ms, H310(RAID0): 150-300ms, H710(Write Back): 150-300ms, H710(Write Through):200-400ms. So I'm sure it's only H310 in JBOD mode does not works as expected. Not related to firmware of hard drives. | 11:42 |
| *** mmcardle has quit IRC | 11:43 | |
| *** mmcardle has joined #openstack-swift | 11:43 | |
| *** kevinc_ has quit IRC | 11:44 | |
| *** bach has joined #openstack-swift | 11:44 | |
| *** mmcardle has quit IRC | 11:48 | |
| *** Honghui has quit IRC | 11:53 | |
| *** ppai has left #openstack-swift | 12:07 | |
| *** bach has quit IRC | 12:11 | |
| *** bach has joined #openstack-swift | 12:11 | |
| *** joeljwright has joined #openstack-swift | 12:14 | |
| *** psharma has quit IRC | 12:16 | |
| *** joeljwright has quit IRC | 12:19 | |
| *** foexle has quit IRC | 12:31 | |
| *** nshaikh has quit IRC | 12:36 | |
| *** nshaikh has joined #openstack-swift | 12:41 | |
| *** tdasilva has left #openstack-swift | 12:42 | |
| *** joeljwright has joined #openstack-swift | 12:45 | |
| *** joeljwright1 has joined #openstack-swift | 12:46 | |
| *** mmcardle has joined #openstack-swift | 12:46 | |
| *** joeljwright has quit IRC | 12:49 | |
| *** nosnos has quit IRC | 12:53 | |
| *** nacim has joined #openstack-swift | 12:54 | |
| -openstackstatus- NOTICE: Zuul is stuck due to earlier networking issues with Gerrit server, work in progress. | 13:03 | |
| *** ChanServ changes topic to "Zuul is stuck due to earlier networking issues with Gerrit server, work in progress." | 13:03 | |
| *** cschwede has quit IRC | 13:11 | |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Restrict keystone cross-tenant ACLs to IDs https://review.openstack.org/86430 | 13:11 |
| *** ChanServ changes topic to "Current Swift Release: 1.13.1 | Priority Reviews: https://wiki.openstack.org/wiki/Swift/PriorityReviews | Channel Logs: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/" | 13:12 | |
| -openstackstatus- NOTICE: Zuul is processing changes now; some results were lost. Use "recheck bug 1317089" if needed. | 13:12 | |
| *** NM1 has joined #openstack-swift | 13:18 | |
| *** mlipchuk has quit IRC | 13:18 | |
| *** mmcardle has quit IRC | 13:19 | |
| *** mmcardle has joined #openstack-swift | 13:20 | |
| *** zul has joined #openstack-swift | 13:22 | |
| *** cschwede has joined #openstack-swift | 13:26 | |
| *** mlipchuk has joined #openstack-swift | 13:27 | |
| *** matsuhashi has quit IRC | 13:35 | |
| *** mmcardle has quit IRC | 13:36 | |
| *** zul has quit IRC | 13:36 | |
| *** zul has joined #openstack-swift | 13:36 | |
| *** joeljwright has joined #openstack-swift | 13:37 | |
| *** joeljwright1 has quit IRC | 13:41 | |
| *** krtaylor_ has joined #openstack-swift | 13:41 | |
| *** krtaylor_ has quit IRC | 13:42 | |
| *** krtaylor has quit IRC | 13:42 | |
| *** NM1 has quit IRC | 13:48 | |
| *** mmcardle has joined #openstack-swift | 13:51 | |
| *** NM1 has joined #openstack-swift | 13:52 | |
| *** krtaylor has joined #openstack-swift | 13:52 | |
| *** foexle has joined #openstack-swift | 13:52 | |
| *** NM1 has quit IRC | 13:53 | |
| *** NM1 has joined #openstack-swift | 14:02 | |
| *** NM1 has quit IRC | 14:04 | |
| *** mmcardle1 has joined #openstack-swift | 14:11 | |
| *** mmcardle has quit IRC | 14:13 | |
| *** NM1 has joined #openstack-swift | 14:14 | |
| *** NM1 has quit IRC | 14:14 | |
| *** shakamunyi has joined #openstack-swift | 14:15 | |
| openstackgerrit | Alex Gaynor proposed a change to openstack/python-swiftclient: Use the six name for importing configparser https://review.openstack.org/92593 | 14:24 |
| *** judd7 has joined #openstack-swift | 14:24 | |
| *** zaitcev has joined #openstack-swift | 14:24 | |
| *** ChanServ sets mode: +v zaitcev | 14:24 | |
| openstackgerrit | Alex Gaynor proposed a change to openstack/python-swiftclient: Make the function tests Python3-import friendly https://review.openstack.org/92593 | 14:28 |
| *** tdasilva has joined #openstack-swift | 14:30 | |
| *** NM1 has joined #openstack-swift | 14:31 | |
| *** mrsnivvel has joined #openstack-swift | 14:32 | |
| *** zul has quit IRC | 14:33 | |
| *** bvandenh has joined #openstack-swift | 14:36 | |
| *** zul has joined #openstack-swift | 14:37 | |
| *** zul has quit IRC | 14:37 | |
| *** Midnightmyth has joined #openstack-swift | 14:38 | |
| *** zul has joined #openstack-swift | 14:38 | |
| notmyname | HP offers that cloud thing everyone is talking about https://www.youtube.com/watch?v=9ntPxdWAWq8 | 14:41 |
| creiht | uhoh | 14:42 |
| creiht | hah | 14:43 |
| portante | oh my | 14:44 |
| portante | 1000 | 14:45 |
| physcx | that is a lot of technology | 14:46 |
| tdasilva | hilarious | 14:48 |
| creiht | notmyname: that's kinda old though isn't it? | 14:48 |
| *** changbl has quit IRC | 14:49 | |
| *** mkollaro has quit IRC | 14:51 | |
| *** joeljwright has quit IRC | 14:54 | |
| *** NM1 has quit IRC | 14:55 | |
| *** NM1 has joined #openstack-swift | 15:01 | |
| *** joeljwright has joined #openstack-swift | 15:02 | |
| *** mkollaro has joined #openstack-swift | 15:09 | |
| *** kevinc_ has joined #openstack-swift | 15:20 | |
| *** shakamunyi has quit IRC | 15:20 | |
| *** shakamunyi has joined #openstack-swift | 15:23 | |
| *** bvandenh has quit IRC | 15:29 | |
| *** haomaiwa_ has quit IRC | 15:29 | |
| notmyname | reminder of today's swift team meeting at 1900utc https://wiki.openstack.org/wiki/Meetings/Swift | 15:32 |
| *** gyee has joined #openstack-swift | 15:40 | |
| *** jasondotstar has joined #openstack-swift | 15:41 | |
| *** nacim has quit IRC | 15:52 | |
| *** nshaikh has quit IRC | 15:58 | |
| *** kevinc_ has quit IRC | 16:01 | |
| *** mwstorer has joined #openstack-swift | 16:01 | |
| *** chandan_kumar has quit IRC | 16:03 | |
| *** praveenkumar has quit IRC | 16:06 | |
| *** kevinc_ has joined #openstack-swift | 16:09 | |
| *** elambert has joined #openstack-swift | 16:09 | |
| openstackgerrit | Alistair Coles proposed a change to openstack/swift: Restrict keystone cross-tenant ACLs to IDs https://review.openstack.org/86430 | 16:40 |
| *** chandan_kumar has joined #openstack-swift | 16:54 | |
| *** mrsnivvel has quit IRC | 16:58 | |
| *** Midnightmyth has quit IRC | 17:04 | |
| *** bach_ has joined #openstack-swift | 17:13 | |
| *** bach has quit IRC | 17:13 | |
| *** russell_h has joined #openstack-swift | 17:14 | |
| *** russell_h has quit IRC | 17:15 | |
| *** russell_h has joined #openstack-swift | 17:15 | |
| *** Midnightmyth has joined #openstack-swift | 17:24 | |
| *** elambert has quit IRC | 17:31 | |
| *** zul has quit IRC | 17:32 | |
| *** krtaylor has quit IRC | 17:33 | |
| *** mlipchuk has quit IRC | 17:41 | |
| openstackgerrit | Thiago da Silva proposed a change to openstack/swift: refactoring functional tests https://review.openstack.org/92643 | 17:42 |
| tdasilva | notmyname: ^^^ I'd to take a stab at re-factoring the functional tests to improve maintainability. tests.py is just really long and I felt we could break it down into smaller chunks | 17:42 |
| *** Midnightmyth has quit IRC | 17:42 | |
| *** elambert has joined #openstack-swift | 17:44 | |
| *** acoles is now known as acoles_away | 17:44 | |
| *** bach_ has quit IRC | 17:44 | |
| *** bach has joined #openstack-swift | 17:45 | |
| *** mrsnivvel has joined #openstack-swift | 17:45 | |
| *** mmcardle1 has quit IRC | 17:46 | |
| *** joeljwright has quit IRC | 17:47 | |
| *** shri has joined #openstack-swift | 18:17 | |
| *** krtaylor has joined #openstack-swift | 18:24 | |
| *** zul has joined #openstack-swift | 18:25 | |
| *** bvandenh has joined #openstack-swift | 18:27 | |
| *** bmac423 has joined #openstack-swift | 18:27 | |
| *** gvernik has joined #openstack-swift | 18:29 | |
| *** kevinc_ has quit IRC | 18:30 | |
| *** notmyname has quit IRC | 18:31 | |
| bmac423 | Can anyone help with swift-informant? I've got keystone and the proxy-server running through mod_wsgi and when I add informant to the pipeline, I get this error when uploading objects: | 18:32 |
| bmac423 | http://paste.openstack.org/show/79445/ | 18:32 |
| *** notmyname_ has joined #openstack-swift | 18:32 | |
| *** ChanServ sets mode: +v notmyname_ | 18:32 | |
| bmac423 | Could be a problem with the swift client, I guess | 18:32 |
| *** notmyname_ is now known as notmyname | 18:32 | |
| zaitcev | bmac423: a quick workaround may be running the client on py3 | 18:36 |
| *** bvandenh has quit IRC | 18:36 | |
| bmac423 | Thanks zaitcev, I actually just got around that error by upgrading the swift client, but I'm still not seeing metrics in graphite... hmmm... | 18:37 |
| pandemicsyn | bmac423: are they hitting statsd ? | 18:41 |
| bmac423 | I don't believe so. is there a good way to tell if they're not showing in graphite? | 18:41 |
| bmac423 | tcpdump is quiet | 18:41 |
| pandemicsyn | tcpdump or statsd in debug mode | 18:41 |
| pandemicsyn | heh | 18:41 |
| *** mkollaro has quit IRC | 18:43 | |
| pandemicsyn | i didn't even know you could run the proxy via mod_wsgi, but i guess you could maybe add a log line/print in https://github.com/pandemicsyn/swift-informant/blob/master/informant/middleware.py#L49 and see if the event is actually getting fired | 18:44 |
| *** mkollaro has joined #openstack-swift | 18:44 | |
| pandemicsyn | oh, also make sure informant isn't at the back of the pipeline or the like | 18:46 |
| bmac423 | pandemicsyn: informant is at the front of the pipeline... I get noise on statsd via tcpdump with manual metrics... (i.e. echo "foo:1|c" | nc -u 127.0.0.1 8125), but nothing from informant | 18:48 |
| *** donagh has joined #openstack-swift | 18:50 | |
| *** jamie_h has quit IRC | 18:51 | |
| bmac423 | pandemicsyn: here's my proxy-server.conf | 18:52 |
| bmac423 | http://paste.openstack.org/show/79457/ | 18:52 |
| bmac423 | manual metrics work on locally on the statsd host and remotely | 18:53 |
| bmac423 | is there any logging associated with swift-informant? | 18:54 |
| *** Kota has joined #openstack-swift | 18:55 | |
| pandemicsyn | yea it'll usually log errors when it hits exceptions and the like | 18:59 |
| bmac423 | syslog? | 18:59 |
| pandemicsyn | yep, where ever you're funneling your proxy logs too | 18:59 |
| *** mjseger has joined #openstack-swift | 19:00 | |
| notmyname | meeting time in #openstack-meeting | 19:00 |
| *** mjseger has quit IRC | 19:01 | |
| *** acoles_away has quit IRC | 19:02 | |
| *** BAKfr has joined #openstack-swift | 19:03 | |
| *** judd7_ has joined #openstack-swift | 19:03 | |
| bmac423 | pandemicsyn: hrmmm... nothing of interest in the logs | 19:06 |
| bmac423 | ran test script... appears all good... Ran 15 tests in 0.012s OK | 19:08 |
| *** judd7_ has quit IRC | 19:08 | |
| *** lpabon has joined #openstack-swift | 19:10 | |
| openstackgerrit | Morgan Fainberg proposed a change to openstack/swift: Add mailmap entry https://review.openstack.org/92671 | 19:15 |
| *** kevinc_ has joined #openstack-swift | 19:16 | |
| bmac423 | rest calls are showing as they should in logs, but nothing going to statsd | 19:18 |
| *** zul has quit IRC | 19:22 | |
| bmac423 | is there anything I may have missed in the swift-informant configuration? perhaps in setup.py? | 19:23 |
| *** occupant has quit IRC | 19:28 | |
| *** zul has joined #openstack-swift | 19:28 | |
| briancline | question for all the performance grand wizards in the house | 19:31 |
| glange | that is quite the title "performance grand wizard" | 19:32 |
| briancline | if I want to experiment with rsync and replication settings for optimal values in my environment, is swift-dispersion-report a good-enough tool to determine the speed at which an object becomes fully consistent after PUT/POST/DELETE? | 19:32 |
| briancline | speed being seconds, I should say | 19:33 |
| briancline | I know it's designed for determining health by trying to hit all partitions, but wasn't sure if there's a better tool for the above purpose | 19:34 |
| *** bmac423 has quit IRC | 19:35 | |
| *** keving has joined #openstack-swift | 19:41 | |
| *** Midnightmyth has joined #openstack-swift | 19:42 | |
| *** nacim has joined #openstack-swift | 19:45 | |
| *** tsg has joined #openstack-swift | 19:46 | |
| briancline | from my understanding a `swift-recon -r object` refers to time for replication runs | 19:46 |
| *** gvernik has quit IRC | 19:55 | |
| *** donagh has left #openstack-swift | 19:56 | |
| *** Kota has quit IRC | 19:57 | |
| *** lpabon has quit IRC | 19:59 | |
| *** ctennis has quit IRC | 20:02 | |
| *** ctennis has joined #openstack-swift | 20:02 | |
| notmyname | this seems like a cool tool: https://github.com/tobert/pcstat | 20:09 |
| notmyname | peluse_: portante: ^^ | 20:09 |
| *** tdasilva has left #openstack-swift | 20:10 | |
| notmyname | mjseger and acoles would like it too, i think. but they aren't here | 20:10 |
| *** tsg has quit IRC | 20:10 | |
| * portante reads | 20:12 | |
| *** joeljwright has joined #openstack-swift | 20:12 | |
| notmyname | portante: for context, it was written in the cassandra community to find things that were using the cache and things that should be using cache. http://informationsurvival.blogspot.com/2014/05/measuring-page-cache-utilization-in.html | 20:13 |
| notmyname | and it's written in go, so you know pandemicsyn and gholt and scotticus will love it ;-) | 20:14 |
| gholt | Yeah, redbo hates go though. | 20:16 |
| redbo | I'm still split on go. I like having a compiled garbage collected system programming language, but you'd think a whole new language wouldn't have so many ugly bits already. | 20:18 |
| redbo | and don't say D, because no matter how hard Walter Bright tries, it's never going to be a thing. | 20:20 |
| redbo | even though pandemicsyn wants the D | 20:21 |
| pandemicsyn | ಠ_ಠ | 20:22 |
| *** ams0 has joined #openstack-swift | 20:22 | |
| *** bach has quit IRC | 20:26 | |
| *** shakamunyi has quit IRC | 20:27 | |
| *** bmac423 has joined #openstack-swift | 20:29 | |
| *** bach has joined #openstack-swift | 20:43 | |
| briancline | any thoughts on the questions from earlier? | 20:44 |
| notmyname | briancline: are you looking for a specific object or the whole cluster? | 20:44 |
| *** Midnightmyth has quit IRC | 20:46 | |
| *** ams0_ has joined #openstack-swift | 20:54 | |
| *** bach has quit IRC | 20:55 | |
| *** ams0 has quit IRC | 20:55 | |
| *** shakamunyi has joined #openstack-swift | 20:55 | |
| bmac423 | pandemicsyn: I'm really racking my brain on this one... any other ideas? Thanks for your help on this. I'll buy you a beer next week if you're at the summit | 20:56 |
| scotticus | notmyname, why would i love it? not a huge fan of go...yet. | 20:56 |
| notmyname | scotticus: I thought you were off writing new stuff in go now :-) | 20:56 |
| scotticus | hah, no time for that! | 20:56 |
| pandemicsyn | bmac423: not really , id try hacking in a debug log line to see if the events are actually making it out | 20:57 |
| *** bach has joined #openstack-swift | 20:58 | |
| bmac423 | Yeah, I don't see much in the logs, so I don't think they're making it out | 20:59 |
| bmac423 | I can upload, delete, list, stat, etc.. but nothing from swift-informant | 20:59 |
| *** jasondotstar has quit IRC | 21:01 | |
| clayg | briancline: i'd guess the dispersion report will be too fine grained to get a good measurement | 21:01 |
| clayg | briancline: unless you're artifically forcing a bunch of parts to move around with a rebalance or something | 21:02 |
| clayg | briancline: maybe you can just delete a partition and instrument timing in the replicator (there may already be some statsd emmissions that would work) | 21:03 |
| pandemicsyn | bmac423: i can't really really think of anything else, the most common issue is usually people typo'ing their statsd ip/port or putting informant at the wrong end of the pipeline | 21:04 |
| redbo | dispersion reports do take quite a while to run. especially a 100% report. | 21:04 |
| pandemicsyn | bmac423: especially since your not see'ing it blow up in the proxy error logs or anything | 21:05 |
| bmac423 | yeah, it's weird | 21:05 |
| *** shakamunyi has quit IRC | 21:07 | |
| *** krtaylor has quit IRC | 21:08 | |
| *** bach has quit IRC | 21:10 | |
| *** ams0_ has quit IRC | 21:10 | |
| *** ams0 has joined #openstack-swift | 21:11 | |
| *** foexle has quit IRC | 21:13 | |
| *** bach has joined #openstack-swift | 21:13 | |
| *** bach has quit IRC | 21:15 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update container storage_policy_index when auditing https://review.openstack.org/83093 | 21:15 |
| clayg | notmyname: what was the DL on that session purposed for "best responses doesn't always return the best response"? | 21:17 |
| notmyname | clayg: http://summit.openstack.org/cfp/details/291 | 21:18 |
| *** bach has joined #openstack-swift | 21:22 | |
| notmyname | clayg: looks interesting, and I hope there is some discussion on it, but the proposed discussion (how much cleverness to configure in best_response) seems like something that needs to happen after the other framework changes happen (ie storage policies) | 21:25 |
| notmyname | IMO. and that's why I didn't choose it as one of the 8 scheduled slots | 21:26 |
| *** bach has quit IRC | 21:26 | |
| *** bach has joined #openstack-swift | 21:27 | |
| clayg | notmyname: no worries, i was wondering if the submitter was talking specifically about DELETE in the face of "some" 404's and it looks like that was one of the things he mentioned. | 21:28 |
| *** joeljwright has quit IRC | 21:29 | |
| notmyname | seems that a storage node returning something different than 404 to the proxy when there is a tombstone could be very useful | 21:29 |
| *** bmac423 has quit IRC | 21:32 | |
| *** NM1 has quit IRC | 21:37 | |
| *** chandan_kumar has quit IRC | 21:37 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add probe tests for Account Reaper https://review.openstack.org/91732 | 21:39 |
| *** yuan has quit IRC | 21:41 | |
| *** yuan has joined #openstack-swift | 21:41 | |
| *** shakamunyi has joined #openstack-swift | 21:41 | |
| peluse_ | notmyname: any time yet to look at the master merge to ec? | 21:45 |
| notmyname | peluse_: nope, sorry | 21:46 |
| *** bach has quit IRC | 21:47 | |
| *** bach has joined #openstack-swift | 21:47 | |
| *** zul has quit IRC | 22:04 | |
| *** bach has quit IRC | 22:12 | |
| *** bach has joined #openstack-swift | 22:24 | |
| *** mkollaro has quit IRC | 22:27 | |
| *** bach has quit IRC | 22:38 | |
| *** bach has joined #openstack-swift | 22:40 | |
| *** bach has quit IRC | 22:40 | |
| *** bach has joined #openstack-swift | 22:41 | |
| *** ophuk has joined #openstack-swift | 22:43 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Fix cross policy versioning DELETE https://review.openstack.org/91503 | 22:45 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Additional functional tests for Storage Policies https://review.openstack.org/91502 | 22:45 |
| *** mrsnivvel has quit IRC | 22:47 | |
| *** shakamunyi has quit IRC | 22:51 | |
| zaitcev | I forgot to bring it up, but users insist we do something about port 6000. | 22:55 |
| zaitcev | The 6200 came up as replacement. | 22:55 |
| pandemicsyn | do users know they can change the config ? | 22:56 |
| clayg | heh | 22:57 |
| pandemicsyn | it is a pita when you tcpdump and wireshark tries to decode the traffic ;) | 22:57 |
| *** bach has quit IRC | 22:59 | |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add probetest for container sync https://review.openstack.org/91685 | 23:02 |
| *** bach has joined #openstack-swift | 23:10 | |
| *** bach has quit IRC | 23:15 | |
| briancline | notmyname: clayg: sorry guys, forcefully pulled from my desk. i'd say I'd be running it against a specific object, but would do a series of different (known) objects just to get a good average. | 23:20 |
| * notmyname is imagining briancline being bodily dragged from his desk. "Noooo, you'll never make me talk!" | 23:21 | |
| briancline | and to hit more than just the one partition | 23:21 |
| briancline | ha, that's about how it usually goes | 23:21 |
| briancline | sometimes there's a telephone involved | 23:21 |
| notmyname | well, the dispersion report is a report of a series of known objects. it's a sample of the partition space rather than a specific partition, though | 23:22 |
| notmyname | so if you run the dispersion report for 100% coverage (which isn't a bad idea), you know what's going on | 23:23 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Keep status_changed_at up-to-date with status changes. https://review.openstack.org/86714 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend direct_client https://review.openstack.org/88144 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add reconciler probetest outline https://review.openstack.org/90934 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policies to container.backend https://review.openstack.org/90935 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Update FakeLogger with better log level handling https://review.openstack.org/90932 | 23:24 |
| clayg | sorry for the spam! | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add container-reconciler daemon https://review.openstack.org/85908 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Make FakeRing a use more real code https://review.openstack.org/90933 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add LRUCache to common.utils https://review.openstack.org/90938 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Extend interface on InternalClient https://review.openstack.org/88152 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Add Storage Policy to container put/delete object https://review.openstack.org/90936 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Enqueue misplaced objects during container replication https://review.openstack.org/90937 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Merge container storage_policy_index values. https://review.openstack.org/86720 | 23:24 |
| openstackgerrit | Clay Gerrard proposed a change to openstack/swift: Expose container info on deleted containers. https://review.openstack.org/86708 | 23:24 |
| briancline | a sample is fine, I think. the key metric I'm looking for is the shortest, avg, longest amount of time it takes to reach 100% consistency (and maybe an stdev later, since it might help show if the results are skewed by a specific set of disks or node(s)) | 23:25 |
| briancline | clayg: oh geeze, show-off | 23:25 |
| notmyname | briancline: it's not that impressive. he just rebases a patch chain about 3x per day | 23:26 |
| notmyname | briancline: that sounds like a statistical analysis of the replication time on each storage node | 23:26 |
| briancline | would I be better served to write up something that uses a known list of objects, PUTs/POSTs/DELETEs accordingly, and hits the relevant storage nodes every ~0.25 - 1sec until it sees a consistent set? | 23:28 |
| briancline | it'll have access to ring data, so it'll know precisely which object servers to hit | 23:28 |
| briancline | (by hit I mean a GET, sorry) | 23:29 |
| briancline | or a HEAD | 23:30 |
| *** ams0 has quit IRC | 23:42 | |
| *** ams0 has joined #openstack-swift | 23:43 | |
| *** ams0 has quit IRC | 23:44 | |
| *** bach has joined #openstack-swift | 23:45 | |
| *** ams0 has joined #openstack-swift | 23:45 | |
| *** ams0 has quit IRC | 23:45 | |
| *** ams0 has joined #openstack-swift | 23:46 | |
| *** ams0 has quit IRC | 23:47 | |
| *** ams0 has joined #openstack-swift | 23:47 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!