| *** frank_young has joined #openstack-swift | 00:20 | |
| *** frank_young has quit IRC | 00:24 | |
| *** tovin07_ has joined #openstack-swift | 00:33 | |
| *** kiennt has joined #openstack-swift | 00:35 | |
| *** frank_young has joined #openstack-swift | 01:33 | |
| *** frank_young has quit IRC | 01:36 | |
| *** frank_young has joined #openstack-swift | 01:36 | |
| *** itlinux_ has quit IRC | 02:00 | |
| *** tovin07__ has joined #openstack-swift | 02:04 | |
| *** tovin07_ has quit IRC | 02:06 | |
| *** kiennt has quit IRC | 02:07 | |
| *** itlinux has joined #openstack-swift | 02:10 | |
| *** OrionX has joined #openstack-swift | 02:25 | |
| *** OrionX has quit IRC | 02:39 | |
| mattoliverau | Morning | 02:39 |
|---|---|---|
| openstackgerrit | Tovin Seven proposed openstack/swift master: OSprofiler in OpenStack Swift https://review.openstack.org/468316 | 02:40 |
| *** OrionX has joined #openstack-swift | 02:49 | |
| *** RayLei has joined #openstack-swift | 03:05 | |
| *** RayLei has quit IRC | 03:05 | |
| *** OrionX has quit IRC | 03:06 | |
| *** tovin07 has joined #openstack-swift | 03:11 | |
| *** tovin07__ has quit IRC | 03:13 | |
| *** kiennt has joined #openstack-swift | 03:18 | |
| *** Sukhdev has quit IRC | 03:20 | |
| *** itlinux has quit IRC | 03:23 | |
| *** itlinux has joined #openstack-swift | 03:30 | |
| *** RayLei has joined #openstack-swift | 03:32 | |
| *** gkadam has joined #openstack-swift | 03:39 | |
| mahatic | good morning | 03:40 |
| mahatic | mattoliverau: o/ | 03:40 |
| *** links has joined #openstack-swift | 03:44 | |
| *** kiennt has quit IRC | 03:52 | |
| *** kiennt has joined #openstack-swift | 03:59 | |
| *** klrmn has quit IRC | 04:00 | |
| *** ChubYann has quit IRC | 04:04 | |
| *** hoonetorg has quit IRC | 04:05 | |
| RayLei | Swauth how to get token from memcached? I do `memcached-tool 127.0.0.1: 11211 dump` seems not to see my token, but can be seen using temauth | 04:11 |
| RayLei | good morning | 04:12 |
| *** ChubYann has joined #openstack-swift | 04:17 | |
| *** psachin has joined #openstack-swift | 04:18 | |
| *** Dinesh_Bhor has joined #openstack-swift | 04:46 | |
| *** Sukhdev has joined #openstack-swift | 04:48 | |
| *** frank_young has quit IRC | 05:03 | |
| *** frank_young has joined #openstack-swift | 05:03 | |
| *** OrionX has joined #openstack-swift | 05:06 | |
| *** tonanhngo has quit IRC | 05:10 | |
| *** OrionX has quit IRC | 05:10 | |
| *** tonanhngo has joined #openstack-swift | 05:16 | |
| *** frank_young has quit IRC | 05:18 | |
| *** frank_young has joined #openstack-swift | 05:28 | |
| *** tonanhngo has quit IRC | 05:30 | |
| *** frank_young has quit IRC | 05:33 | |
| *** cshastri has joined #openstack-swift | 05:37 | |
| *** frank_young has joined #openstack-swift | 05:43 | |
| *** kiennt has quit IRC | 05:45 | |
| *** winggundamth has joined #openstack-swift | 05:53 | |
| openstackgerrit | junbo.li proposed openstack/swift master: [Api-ref] fix response status https://review.openstack.org/484254 | 05:54 |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Add multiple worker processes strategy to reconstructor https://review.openstack.org/478416 | 05:54 |
| *** rcernin_ has joined #openstack-swift | 06:00 | |
| *** kiennt has joined #openstack-swift | 06:03 | |
| openstackgerrit | Clay Gerrard proposed openstack/swift master: Add multiple worker processes strategy to reconstructor https://review.openstack.org/478416 | 06:07 |
| *** kiennt_ has joined #openstack-swift | 06:17 | |
| *** kiennt has quit IRC | 06:17 | |
| *** SkyRocknRoll has joined #openstack-swift | 06:22 | |
| *** cschwede_ has joined #openstack-swift | 06:23 | |
| *** ChubYann has quit IRC | 06:24 | |
| *** Sukhdev has quit IRC | 06:50 | |
| *** itlinux has quit IRC | 06:59 | |
| *** itlinux has joined #openstack-swift | 07:00 | |
| *** tonanhngo has joined #openstack-swift | 07:02 | |
| *** tonanhngo has quit IRC | 07:03 | |
| *** itlinux has quit IRC | 07:08 | |
| *** dhellmann has quit IRC | 07:16 | |
| *** dhellmann has joined #openstack-swift | 07:17 | |
| *** geaaru has joined #openstack-swift | 07:30 | |
| *** oshritf has joined #openstack-swift | 07:30 | |
| *** tesseract has joined #openstack-swift | 07:32 | |
| *** cbartz has joined #openstack-swift | 07:36 | |
| acoles | good morning | 07:49 |
| acoles | kota_: cschwede_ if you have chance to review patch 477000 I would really appreciate your feedback | 07:53 |
| patchbot | https://review.openstack.org/#/c/477000/ - swift - Ring rebalance respects co-builders' last_part_moves | 07:53 |
| *** cbartz has quit IRC | 07:54 | |
| acoles | I'm looking at patch 478416 today | 08:03 |
| patchbot | https://review.openstack.org/#/c/478416/ - swift - Add multiple worker processes strategy to reconstr... | 08:03 |
| cschwede_ | acoles: alright, will review it today | 08:22 |
| acoles | cschwede_: thanks | 08:27 |
| *** tesseract has quit IRC | 08:50 | |
| *** tesseract has joined #openstack-swift | 08:53 | |
| *** oshritf has quit IRC | 08:56 | |
| kiennt_ | cschwede_: Hi, i want to ask about multinode gate job for rolling upgrade. I asked in channel last week and they said I should ask you. | 09:00 |
| kiennt_ | Thank you | 09:00 |
| cschwede_ | kiennt_: Hi! how can i help? | 09:05 |
| kiennt_ | cschwede_: hi, i am working on Heat rolling upgrade and Heat core team asked me setup multinode gate job for rolling-upgrade testing. | 09:06 |
| kiennt_ | Could you help me to find how to setup it? Because Swift already has assert:support-rolling-upgrade | 09:08 |
| kiennt_ | How Swift setup multinode gate? | 09:08 |
| cschwede_ | kiennt_: I'm not sure about the criteria for that flag. Swift rolling upgrades are not tested yet in the gate, but Swift supports that since the beginning (afaict) | 09:11 |
| cschwede_ | kiennt_: I proposed a patch for the gate sometime ago: https://review.openstack.org/#/c/304465/ | 09:11 |
| patchbot | patch 304465 - openstack-infra/devstack-gate - Use subnodes for Swift storage nodes in a multinod... | 09:11 |
| cschwede_ | kiennt_: but there was not that much feedback, so i stopped working on that | 09:12 |
| kiennt_ | cschwede_: Oh I see, thanks for your help (and your patch too). Very helpful info :0- | 09:15 |
| kiennt_ | :) | 09:15 |
| openstackgerrit | junbo.li proposed openstack/swift master: [Api-ref] fix response status https://review.openstack.org/484254 | 09:22 |
| openstackgerrit | junbo.li proposed openstack/swift master: Docs migration from openstack-manuals https://review.openstack.org/484182 | 09:33 |
| *** oshritf has joined #openstack-swift | 09:46 | |
| *** tovin07 has quit IRC | 09:52 | |
| *** cbartz has joined #openstack-swift | 10:08 | |
| *** jistr is now known as jistr|afk | 10:09 | |
| *** d0ugal has joined #openstack-swift | 10:13 | |
| *** d0ugal has quit IRC | 10:19 | |
| *** mvk has quit IRC | 10:33 | |
| *** kiennt_ has quit IRC | 10:37 | |
| *** logan- has left #openstack-swift | 10:53 | |
| *** frank_young has quit IRC | 10:58 | |
| *** mvk has joined #openstack-swift | 11:00 | |
| *** tonanhngo has joined #openstack-swift | 11:04 | |
| *** tonanhngo has quit IRC | 11:05 | |
| *** frank_young has joined #openstack-swift | 11:08 | |
| *** frank_young has quit IRC | 11:13 | |
| *** frank_young has joined #openstack-swift | 11:24 | |
| *** hoonetorg has joined #openstack-swift | 11:32 | |
| *** logan- has joined #openstack-swift | 11:40 | |
| *** Guest7927 has joined #openstack-swift | 11:43 | |
| Guest7927 | Hi | 11:43 |
| Guest7927 | Is there any other caching service Swift uses, except for memcached? | 11:45 |
| *** logan- has left #openstack-swift | 12:00 | |
| *** OrionX has joined #openstack-swift | 12:07 | |
| *** OrionX has quit IRC | 12:12 | |
| *** jistr|afk is now known as jistr | 12:17 | |
| *** MVenesio has joined #openstack-swift | 12:28 | |
| RayLei | Swauth how to get token from memcached? I do `memcached-tool 127.0.0.1: 11211 dump` seems not to see my token, but can be seen using temauth | 12:33 |
| cbartz | Have you already used your token in a request? It gets stored the first time used. | 12:37 |
| RayLei | Yes, and how long will it last in memcache? How can I see it? | 12:39 |
| RayLei | I know that if the first time to do the authentication user's request, it will only store the token inside the cluster, with token to request the cluster when the resource will be stored in memcache token, but I did not find it | 12:42 |
| cbartz | Time of the token. If not set explicitly, its 86400 seconds | 12:42 |
| RayLei | But I have never seen it in a token_life time | 12:43 |
| RayLei | with `memcached-tool localhost:11211 dump` or swift.common.memcached.MemcacheRing('127.0.0.1:11211').get('AUTH_/user/myaccount:myuser') ,do not see the token | 12:46 |
| cbartz | memcache_key should be: '%s/auth/%s' % (self.reseller_prefix, token) | 12:48 |
| RayLei | you can see here | 12:50 |
| RayLei | swift auth export OS_STORAGE_URL=http://192.168.2.133:8080/v1/AUTH_71743be6-f7e8-408e-97e8-5e85d139f639 export OS_AUTH_TOKEN=AUTH_tk1ef0e8d499724897b957e77379133aaf | 12:50 |
| *** catintheroof has joined #openstack-swift | 12:51 | |
| RayLei | swift auth； export OS_STORAGE_URL=http://192.168.2.133:8080/v1/AUTH_71743be6-f7e8-408e-97e8-5e85d139f639 ；export OS_AUTH_TOKEN=AUTH_tk1ef0e8d499724897b957e77379133aaf | 12:51 |
| RayLei | >>> print mem.get('AUTH_/token/AUTH_tk1ef0e8d499724897b957e77379133aaf') | 12:51 |
| RayLei | The result is NONE | 12:52 |
| cbartz | Try: print mem.get('AUTH_/auth/AUTH_tk1ef0e8d499724897b957e77379133aaf') | 12:54 |
| RayLei | >>> print mem.get('AUTH_/auth/AUTH_tk1ef0e8d499724897b957e77379133aaf') | 12:57 |
| RayLei | [1500344226, u'test:tester,test,AUTH_71743be6-f7e8-408e-97e8-5e85d139f639'] | 12:57 |
| RayLei | cool | 12:58 |
| cbartz | Yeah. You see. You used the wrong memcache key, the one tempauth uses | 12:58 |
| RayLei | Very strange, has not been before, I also used the correct key, now have, interesting | 12:59 |
| RayLei | In short, thank you for your reply | 13:03 |
| cbartz | Your welcome. | 13:05 |
| *** honga has quit IRC | 13:06 | |
| *** RayLei has quit IRC | 13:20 | |
| openstackgerrit | junbo.li proposed openstack/swift master: [Api-ref] fix response status https://review.openstack.org/484371 | 13:25 |
| *** gkadam has quit IRC | 13:26 | |
| *** SkyRocknRoll has quit IRC | 13:27 | |
| openstackgerrit | junbo.li proposed openstack/swift master: [Api-ref] fix response status https://review.openstack.org/484371 | 13:30 |
| openstackgerrit | junbo.li proposed openstack/swift master: [Api-ref] fix response status https://review.openstack.org/484371 | 13:32 |
| *** klamath has joined #openstack-swift | 13:36 | |
| *** links has quit IRC | 13:37 | |
| *** frank_young has quit IRC | 13:38 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Make dict deletion idempotent in dump_recon_cache https://review.openstack.org/484375 | 13:40 |
| *** klamath has quit IRC | 13:58 | |
| *** klamath has joined #openstack-swift | 13:58 | |
| *** tonanhngo has joined #openstack-swift | 14:01 | |
| *** tonanhngo has quit IRC | 14:02 | |
| *** chlong has joined #openstack-swift | 14:05 | |
| *** lucasxu has joined #openstack-swift | 14:06 | |
| *** winggundamth has quit IRC | 14:08 | |
| *** oshritf has quit IRC | 14:09 | |
| *** chlong has quit IRC | 14:10 | |
| *** gyee has joined #openstack-swift | 14:37 | |
| *** cshastri has quit IRC | 14:41 | |
| *** psachin has quit IRC | 14:48 | |
| *** oshritf has joined #openstack-swift | 14:48 | |
| *** klrmn has joined #openstack-swift | 14:50 | |
| *** lucasxu has quit IRC | 15:00 | |
| *** gyee has quit IRC | 15:06 | |
| *** OrionX has joined #openstack-swift | 15:06 | |
| *** gyee has joined #openstack-swift | 15:08 | |
| *** lucasxu has joined #openstack-swift | 15:27 | |
| *** oshritf has quit IRC | 15:30 | |
| *** chlong has joined #openstack-swift | 15:36 | |
| *** chlong has quit IRC | 15:38 | |
| *** MVenesio has quit IRC | 15:42 | |
| *** chlong has joined #openstack-swift | 15:45 | |
| *** hoonetorg has quit IRC | 15:57 | |
| *** cbartz has quit IRC | 16:08 | |
| *** klrmn has quit IRC | 16:11 | |
| *** rcernin_ has quit IRC | 16:19 | |
| *** JimCheung has joined #openstack-swift | 16:30 | |
| *** hoonetorg has joined #openstack-swift | 16:40 | |
| *** spotz has quit IRC | 16:42 | |
| *** nadeem has quit IRC | 16:43 | |
| *** nadeem has joined #openstack-swift | 16:43 | |
| *** spotz has joined #openstack-swift | 16:45 | |
| *** ChubYann has joined #openstack-swift | 16:47 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Remove unnecessary exception handling https://review.openstack.org/484428 | 16:54 |
| *** hoonetorg has quit IRC | 16:54 | |
| *** frank_young has joined #openstack-swift | 16:59 | |
| *** hoonetorg has joined #openstack-swift | 17:03 | |
| *** chlong has quit IRC | 17:03 | |
| *** frank_young has quit IRC | 17:04 | |
| *** mvk has quit IRC | 17:06 | |
| *** klrmn has joined #openstack-swift | 17:08 | |
| *** mattoliverau has quit IRC | 17:11 | |
| *** chlong has joined #openstack-swift | 17:19 | |
| *** mattoliverau has joined #openstack-swift | 17:19 | |
| *** ChanServ sets mode: +v mattoliverau | 17:19 | |
| *** tonanhngo has joined #openstack-swift | 17:28 | |
| *** tonanhngo has quit IRC | 17:32 | |
| *** hoonetorg has quit IRC | 17:34 | |
| *** mvk has joined #openstack-swift | 17:37 | |
| *** frank_young has joined #openstack-swift | 17:37 | |
| *** tonanhngo has joined #openstack-swift | 17:38 | |
| *** MVenesio has joined #openstack-swift | 17:39 | |
| *** frank_young has quit IRC | 17:42 | |
| *** SkyRocknRoll has joined #openstack-swift | 17:53 | |
| *** hoonetorg has joined #openstack-swift | 17:54 | |
| *** hoonetorg has quit IRC | 17:59 | |
| *** tesseract has quit IRC | 18:01 | |
| *** geaaru has quit IRC | 18:21 | |
| *** Sukhdev has joined #openstack-swift | 18:24 | |
| *** rcernin has joined #openstack-swift | 18:26 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Avoid deleting new per disk reconstructor stats when aggregating https://review.openstack.org/484448 | 18:34 |
| notmyname | good morning | 18:35 |
| *** JimCheung has quit IRC | 18:45 | |
| *** JimCheung has joined #openstack-swift | 18:45 | |
| *** Guest7927 has quit IRC | 18:52 | |
| *** SkyRocknRoll has quit IRC | 19:02 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Remove unnecessary exception handling https://review.openstack.org/484428 | 19:05 |
| *** frank_young has joined #openstack-swift | 19:06 | |
| *** Nathaniel has joined #openstack-swift | 19:07 | |
| *** Nathaniel is now known as Guest90042 | 19:07 | |
| Guest90042 | Why swift use memcached? | 19:07 |
| *** klamath has quit IRC | 19:07 | |
| *** klamath has joined #openstack-swift | 19:08 | |
| *** klamath has quit IRC | 19:09 | |
| *** klamath has joined #openstack-swift | 19:09 | |
| *** frank_young has quit IRC | 19:11 | |
| *** JimCheung has quit IRC | 19:12 | |
| *** JimCheung has joined #openstack-swift | 19:12 | |
| clayg | acoles: so i have this bug -> https://bugs.launchpad.net/swift/+bug/1704858 | 19:28 |
| openstack | Launchpad bug 1704858 in OpenStack Object Storage (swift) "auditor recon dump is not consistent on new nodes" [Low,New] | 19:28 |
| clayg | but your fix doesn't really close that bug - since all it does is just make the key not ever show up | 19:28 |
| clayg | which is a useful behavior for the related change - but it means your "fix" isn't any more "obvious" | 19:28 |
| clayg | in this case the question is "*why* are we deleting these keys" in my change the reason *why* is obvious; leaving only the question of "why is the *interface* for deleting keys so bloody strange?" | 19:31 |
| *** OrionX has quit IRC | 19:31 | |
| clayg | and the answer is, it doesn't need be, the interface is just wrong - and fixing the interface (while not fixing the answering the issue in the auditor) won't make anything wrose! | 19:31 |
| *** JimCheung has quit IRC | 19:33 | |
| *** JimCheung has joined #openstack-swift | 19:33 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Make dict deletion idempotent in dump_recon_cache https://review.openstack.org/484375 | 19:37 |
| acoles | clayg: ok ^^ I tried to make the commit message explain why i wasn't fixing the bug per se, because the intent is to fix the utils helper, not to fix weird auditor behavior. | 19:38 |
| *** JimCheung has quit IRC | 19:38 | |
| *** Guest90042 has quit IRC | 19:40 | |
| *** JimCheung has joined #openstack-swift | 19:45 | |
| *** frank_young has joined #openstack-swift | 19:46 | |
| clayg | so this is the patch 484448 | 19:46 |
| patchbot | https://review.openstack.org/#/c/484448/ - swift - Avoid deleting new per disk reconstructor stats wh... | 19:46 |
| acoles | clayg: https://review.openstack.org/#/c/484375 | 19:46 |
| patchbot | patch 484375 - swift - Make dict deletion idempotent in dump_recon_cache | 19:46 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Make dict deletion idempotent in dump_recon_cache https://review.openstack.org/484375 | 19:47 |
| *** JimCheung has quit IRC | 19:49 | |
| *** frank_young has quit IRC | 19:50 | |
| clayg | acoles: ok, i have +2'd patch 484375 | 19:52 |
| patchbot | https://review.openstack.org/#/c/484375/ - swift - Make dict deletion idempotent in dump_recon_cache | 19:52 |
| clayg | i think it's obviously correct and makes swift slightly (but measurably non-zero) better | 19:53 |
| clayg | I'm going to rebase my multi-worker reconstructor work on top of that - and hopefully we can get that fix in w/o too much fuss so I can make use of that interface w/o having to add a bunch of weird work-arounds or unrelated test code | 19:53 |
| clayg | cool cool | 19:53 |
| acoles | clayg: +1 I like not having weird test assertions | 19:54 |
| clayg | acoles: this patch should be squahed patched 484375 | 19:56 |
| acoles | yes | 19:56 |
| *** aj701 has quit IRC | 19:59 | |
| *** JimCheung has joined #openstack-swift | 20:01 | |
| *** aj701 has joined #openstack-swift | 20:04 | |
| *** JimCheung has quit IRC | 20:05 | |
| *** cschwede_ has quit IRC | 20:15 | |
| *** OrionX has joined #openstack-swift | 20:16 | |
| *** chlong has quit IRC | 20:27 | |
| *** chlong has joined #openstack-swift | 20:32 | |
| *** vint_bra has joined #openstack-swift | 20:54 | |
| *** JimCheung has joined #openstack-swift | 20:56 | |
| *** lucasxu has quit IRC | 21:01 | |
| *** MVenesio has quit IRC | 21:12 | |
| *** Sukhdev has quit IRC | 21:26 | |
| *** chlong has quit IRC | 21:26 | |
| *** logan- has joined #openstack-swift | 21:32 | |
| *** JimCheung has quit IRC | 21:32 | |
| tdasilva | zaitcev: do you remember what cause this error: http://eavesdrop.openstack.org/irclogs/%23openstack-swift/%23openstack-swift.2017-05-22.log.html#t2017-05-22T21:02:32 | 21:33 |
| *** logan- has left #openstack-swift | 21:33 | |
| *** rcernin has quit IRC | 21:33 | |
| openstackgerrit | Merged openstack/swift master: Fix redundant UTF-8 check https://review.openstack.org/484107 | 21:34 |
| zaitcev | tdasilva: in my case it was caused by building the stock 1.4.0 at an Intel CPU, and then trying to run the resulting RPM on an AMD CPU. | 21:34 |
| zaitcev | tdasilva: IIRC we merged a fix for that upstream. | 21:35 |
| tdasilva | ok | 21:35 |
| zaitcev | Actually, we merged 2 | 21:35 |
| tdasilva | zaitcev: vint_bra is running into this issue | 21:35 |
| tdasilva | zaitcev: and that was released in 1.5.0? | 21:35 |
| timburke | https://github.com/openstack/liberasurecode/commit/960cdd0 and https://github.com/openstack/liberasurecode/commit/0962144 | 21:36 |
| zaitcev | Fix #1 is to stop using <math.h>. As long as you aren't using floating point, gcc isn't over-optimizing and then we're good | 21:36 |
| zaitcev | Fix #2 is to allow RPM to override CPU detection and force gcc to build a portable code (which is slower, yes). | 21:36 |
| zaitcev | tdasilva: IIRC released in 1.5.0 | 21:37 |
| tdasilva | vint_bra: sounds like the ideal would be for you guys to upgrade to 1.5.0 | 21:38 |
| *** caiobrentano has joined #openstack-swift | 21:42 | |
| *** frank_young has joined #openstack-swift | 21:44 | |
| *** JimCheung has joined #openstack-swift | 21:48 | |
| *** frank_young has quit IRC | 21:48 | |
| vint_bra | tdasilva: Great! It works! :) | 21:50 |
| tdasilva | vint_bra: did you upgrade both to 1.5? | 21:50 |
| vint_bra | yes | 21:51 |
| tdasilva | vint_bra: nice! | 21:51 |
| vint_bra | tdasilva: Thanks for everything! | 21:52 |
| vint_bra | zaitcev: Thanks my friend | 21:52 |
| *** itlinux_ has joined #openstack-swift | 21:52 | |
| tdasilva | vint_bra: :) | 21:53 |
| *** caiobrentano has quit IRC | 21:55 | |
| *** caiobrentano has joined #openstack-swift | 21:56 | |
| *** JimCheung has joined #openstack-swift | 21:59 | |
| *** caiobrentano has quit IRC | 22:01 | |
| *** OrionX_ has joined #openstack-swift | 22:10 | |
| *** vint_bra has quit IRC | 22:13 | |
| openstackgerrit | OpenStack Proposal Bot proposed openstack/swift master: Updated from global requirements https://review.openstack.org/88736 | 22:13 |
| *** OrionX has quit IRC | 22:14 | |
| *** OrionX_ has quit IRC | 22:14 | |
| *** tonanhngo has quit IRC | 22:20 | |
| *** tonanhngo has joined #openstack-swift | 22:22 | |
| *** tonanhngo_ has joined #openstack-swift | 22:26 | |
| *** tonanhngo has quit IRC | 22:26 | |
| *** tonanhngo_ has quit IRC | 22:30 | |
| *** Nathaniel has joined #openstack-swift | 22:31 | |
| Nathaniel | Can I use Swift without memcached? | 22:31 |
| *** Nathaniel is now known as Guest8041 | 22:31 | |
| notmyname | probably (depending on the auth system you're using), but it's not recommended | 22:32 |
| clayg | Nathaniel it depends on the auth system, but generally yet, only you don't want to because of info checks | 22:32 |
| clayg | timburke: thanks for looking at patch 484375 | 22:34 |
| patchbot | https://review.openstack.org/#/c/484375/ - swift - Make dict deletion idempotent in dump_recon_cache | 22:34 |
| clayg | that has got to be the quickest turn around on a cleanup fix up ever! | 22:34 |
| Guest8041 | If so, is there any other auth system that does not use memcached provided by swift? | 22:35 |
| notmyname | Guest8041: I don't want to out-of-hand tell you "just use memcache", but may I ask why using or not is such a big question? | 22:36 |
| notmyname | beyond auth, as clayg said, memcache is very important for caching account and container stats. it vastly improves the speed of all requests in the cluster | 22:37 |
| clayg | it's required - just run it | 22:37 |
| Guest8041 | but memcached causes data duplication between main memory and storage memory. | 22:38 |
| Guest8041 | I'm considering to use swift with non-volatile-memory | 22:38 |
| Guest8041 | so I just wonder that | 22:38 |
| Guest8041 | thank you for answer :) | 22:40 |
| clayg | none of the object data is kept in memcache | 22:41 |
| clayg | it's just for stuff like auth tokens, or the account/container stats tables | 22:41 |
| clayg | the auth token would have to be durable in a auth-database db somewhere (but they probably will need to cache it if you're asking about it every request) | 22:41 |
| *** itlinux_ has quit IRC | 22:42 | |
| Guest8041 | There was a part that I misunderstood. thank you. | 22:42 |
| clayg | the account/container stats tables are durable on disk - but it can be annoying to keep them in the page cache on all the nodes all the time for the entire working set of containers - normally better to warm it up in the memory in the proxy tier - but YMMV | 22:43 |
| Guest8041 | oh.. I see | 22:43 |
| * clayg shrugs | 22:43 | |
| *** Guest8041 has quit IRC | 22:56 | |
| notmyname | doh! he left | 22:57 |
| notmyname | I was going to suggest that if he's got nvram available, using something like Intel's CAS on the object server can provide some nice performance benefit | 22:58 |
| *** JimCheung has quit IRC | 22:59 | |
| *** JimCheung has joined #openstack-swift | 23:00 | |
| *** OrionX has joined #openstack-swift | 23:00 | |
| *** frank_young has joined #openstack-swift | 23:03 | |
| *** OrionX has quit IRC | 23:04 | |
| *** catintheroof has quit IRC | 23:07 | |
| *** frank_young has quit IRC | 23:07 | |
| *** chlong has joined #openstack-swift | 23:24 | |
| *** kei_yama has joined #openstack-swift | 23:26 | |
| *** kei_yama has quit IRC | 23:48 | |
| *** kei_yama has joined #openstack-swift | 23:53 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!