seongsoocho | morning | 00:08 |
---|---|---|
*** threestrands has joined #openstack-swift | 00:35 | |
kota_ | morning | 00:56 |
openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Only add __del__ to HTTPConnection shim on py3 https://review.opendev.org/721051 | 02:08 |
kota_ | thx timburke to resolve the problem! | 03:01 |
*** manuvakery has joined #openstack-swift | 03:01 | |
*** psachin has joined #openstack-swift | 03:37 | |
*** evrardjp has quit IRC | 04:23 | |
*** tkajinam has quit IRC | 05:12 | |
*** tkajinam has joined #openstack-swift | 05:13 | |
*** manuvakery has quit IRC | 05:31 | |
*** evrardjp has joined #openstack-swift | 05:41 | |
openstackgerrit | Merged openstack/python-swiftclient master: Update docs building https://review.opendev.org/721069 | 06:11 |
*** rpittau|afk is now known as rpittau | 07:12 | |
openstackgerrit | Matthew Oliver proposed openstack/swift master: Auto-sharding: Initial steps https://review.opendev.org/667030 | 07:15 |
openstackgerrit | Matthew Oliver proposed openstack/swift master: sharding: first attempt at _elect_leader https://review.opendev.org/667579 | 07:15 |
mattoliverau | just a rebase, as I want to poke at it to see where it's at :) | 07:16 |
*** threestrands_ has joined #openstack-swift | 07:18 | |
*** threestrands_ has quit IRC | 07:18 | |
*** threestrands has quit IRC | 07:21 | |
*** ccamacho has joined #openstack-swift | 08:12 | |
*** psachin has quit IRC | 09:09 | |
*** psachin has joined #openstack-swift | 09:17 | |
*** dtantsur|afk is now known as dtantsur | 09:21 | |
*** gregwork has quit IRC | 09:24 | |
*** rcernin has quit IRC | 09:26 | |
*** rcernin has joined #openstack-swift | 09:51 | |
openstackgerrit | OpenStack Release Bot proposed openstack/swift stable/ussuri: Update .gitreview for stable/ussuri https://review.opendev.org/721214 | 09:59 |
openstackgerrit | OpenStack Release Bot proposed openstack/swift stable/ussuri: Update TOX_CONSTRAINTS_FILE for stable/ussuri https://review.opendev.org/721215 | 10:00 |
openstackgerrit | OpenStack Release Bot proposed openstack/swift master: Update master for stable/ussuri https://review.opendev.org/721216 | 10:00 |
*** rpittau is now known as rpittau|bbl | 10:15 | |
*** tkajinam has quit IRC | 11:21 | |
*** rpittau|bbl is now known as rpittau | 12:03 | |
*** irclogbot_2 has quit IRC | 13:06 | |
*** irclogbot_2 has joined #openstack-swift | 13:08 | |
*** psachin has quit IRC | 13:34 | |
*** hoonetorg has quit IRC | 13:40 | |
*** tkajinam has joined #openstack-swift | 13:53 | |
*** hoonetorg has joined #openstack-swift | 13:59 | |
-openstackstatus- NOTICE: Zuul is temporarily offline; service should be restored in about 15 minutes. | 14:00 | |
*** tkajinam has quit IRC | 14:59 | |
openstackgerrit | Tim Burke proposed openstack/swift master: Update master for stable/ussuri https://review.opendev.org/721216 | 15:05 |
*** gyee has joined #openstack-swift | 15:40 | |
*** hoonetorg has quit IRC | 15:46 | |
-openstackstatus- NOTICE: Gerrit will be restarted to correct a misconfiguration which caused some git mirrors to have outdated references. | 15:48 | |
*** ianychoi_ is now known as ianychoi | 15:53 | |
*** hoonetorg has joined #openstack-swift | 15:59 | |
*** manuvakery has joined #openstack-swift | 16:03 | |
*** rpittau is now known as rpittau|afk | 16:07 | |
*** dtantsur is now known as dtantsur|afk | 16:32 | |
*** evrardjp has quit IRC | 16:35 | |
*** evrardjp has joined #openstack-swift | 16:35 | |
seongsoocho | Hi ! Does anyone tell me a MTU size between proxy and object node? I set it 1500 but there are too many packet retransmission and drop occures.. | 16:50 |
seongsoocho | I'm looking for set MTU to 9000. and other sysctl settings. | 16:51 |
*** abelur has quit IRC | 16:54 | |
*** abelur has joined #openstack-swift | 16:54 | |
openstackgerrit | Tim Burke proposed openstack/python-swiftclient master: Add py38 classifier https://review.opendev.org/721313 | 17:10 |
timburke | seongsoocho, typically we've seen customers running into trouble when MTU is too *high* -- customers would have their servers configured for 9000 but not their switches | 17:23 |
timburke | 1500 was usually fine... are you sure everybody's configured the same way? | 17:23 |
seongsoocho | timburke: I tried to change MTU to 9000 both servers and switches. But I'm not sure It is a best solution. I will find another way to solve my problem ( keep the MTU 1500 ) 🙂 | 17:32 |
*** evrardjp has quit IRC | 17:44 | |
openstackgerrit | Merged openstack/swift stable/ussuri: Update .gitreview for stable/ussuri https://review.opendev.org/721214 | 17:46 |
*** evrardjp has joined #openstack-swift | 17:49 | |
manuvakery | Guys .. am having trouble with connection timeout issues.. when object server tries to update the container db I can see timout happens for http connection (0.5s) or for response (3s) .. is it due to the container servers are not able to handle the incoming traffic .. did not notice high io . Is there any hard limit for container server max connection.. | 18:14 |
*** jrosser has quit IRC | 18:41 | |
*** jrosser has joined #openstack-swift | 18:42 | |
timburke | manuvakery, that's normal, expected behavior -- provided your object-updaters can still clear the async-pendings reasonably quickly. you can crank up the timeout if you like, but there will always come a point where container DBs can't keep up with updates for some period | 18:48 |
timburke | maybe it's too many connections, usually it's not enough iops (if containers are on spinning disks, there's almost *never* enough iops) -- but it's almost always because of a temporary spike, so we can just save the update for later and get to it eventually | 18:51 |
timburke | if it's *not* because of temporary spikes, you probably ought to look into ratelimiting, either per account or per container (or both) | 18:52 |
timburke | when you say you "did not notice high io" are you talking at the disk level (with iostat or similar), or proxy throughput, or proxy requests/sec... something else? | 18:54 |
manuvakery | I meant disk level iostat | 18:55 |
manuvakery | Ok.. if it's normal then fine .. I am using hdd for container db .. may be using an SSD can help I think | 18:59 |
*** ccamacho has quit IRC | 20:01 | |
clayg | rledisez: do you guys have some "bucket inventory" like tools? Rackspace used to use something like https://opendev.org/x/slogging | 20:43 |
rledisez | clayg: nope. but I wanted for a long to write one. never found time to do it :/ | 20:50 |
rledisez | now I can just deploy this one \o/ | 20:50 |
clayg | we'll see it's quite old | 20:50 |
clayg | but so it swift AMIRITE!? | 20:51 |
rledisez | do you know how it works? My thinking was that container-auditor should emit a serie for each feature .increment("versioning-enabled"), etc… | 20:52 |
*** manuvakery has quit IRC | 21:13 | |
*** dosaboy has quit IRC | 21:46 | |
*** dosaboy has joined #openstack-swift | 22:02 | |
*** rcernin has quit IRC | 22:05 | |
*** rcernin has joined #openstack-swift | 22:06 | |
DHE | I have found a problem with using the ring composer... the one-partition-moved-at-a-time rule does not apply the same way any more... | 22:13 |
*** tkajinam has joined #openstack-swift | 22:43 | |
openstackgerrit | Merged openstack/swift master: Update master for stable/ussuri https://review.opendev.org/721216 | 22:52 |
mattoliverau | morning | 23:09 |
*** threestrands has joined #openstack-swift | 23:18 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!