| zaitcev | acoles: you around? How is this a good thing: def __hash__(self): return 1 | 00:38 |
|---|---|---|
| *** timburke__ has quit IRC | 00:47 | |
| *** timburke__ has joined #openstack-swift | 00:48 | |
| *** ChanServ sets mode: +v timburke__ | 00:48 | |
| *** rcernin has quit IRC | 01:34 | |
| *** baojg has joined #openstack-swift | 01:34 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift master: relinker: Add policy to relinker progress output https://review.opendev.org/c/openstack/swift/+/780440 | 01:39 |
| openstackgerrit | Matthew Oliver proposed openstack/swift master: relinker: Add /recon/relinker endpoint and drop progress stats https://review.opendev.org/c/openstack/swift/+/779655 | 01:39 |
| *** rcernin has joined #openstack-swift | 01:45 | |
| *** rcernin has quit IRC | 01:46 | |
| *** rcernin has joined #openstack-swift | 01:46 | |
| seongsoocho | yay ! I just finished my first partition power increase job! It took almost 6 weeks. ... Thank you for making such a fantastic feature. | 02:44 |
| *** rcernin has quit IRC | 02:45 | |
| *** rcernin has joined #openstack-swift | 02:57 | |
| *** rcernin has quit IRC | 02:57 | |
| *** rcernin has joined #openstack-swift | 02:58 | |
| zaitcev | seongsoocho: I cannot believe it took that long. Can you give me an order of magnitude for the number of objects in the cluster (I understand that exact number may be proprietary)? | 02:59 |
| seongsoocho | zaitcev: 1,025,311,000 object with 480 disks. Looking at the relink log, it says that 10 million diskfiles were processed per disk. | 03:03 |
| zaitcev | seongsoocho: thanks | 03:03 |
| zaitcev | well the relinker operates on directories which is slow by necessity | 03:04 |
| seongsoocho | yes.. I added additional log to know how many diskfiles were processed per 1 min, the report said about 2000 diskfiles were process per 1min. | 03:06 |
| *** rcernin has quit IRC | 03:31 | |
| *** evrardjp has quit IRC | 03:33 | |
| *** evrardjp has joined #openstack-swift | 03:33 | |
| *** rcernin has joined #openstack-swift | 03:38 | |
| *** rcernin has quit IRC | 03:52 | |
| *** rcernin has joined #openstack-swift | 03:52 | |
| *** rcernin has quit IRC | 03:53 | |
| *** rcernin has joined #openstack-swift | 03:54 | |
| openstackgerrit | Matthew Oliver proposed openstack/swift master: Stop pushing acceptor and donor SR to shrinking shards https://review.opendev.org/c/openstack/swift/+/771908 | 04:41 |
| *** psachin has joined #openstack-swift | 05:53 | |
| *** psachin has quit IRC | 06:01 | |
| *** baojg has quit IRC | 06:43 | |
| *** baojg has joined #openstack-swift | 06:44 | |
| *** m75abrams has joined #openstack-swift | 06:50 | |
| *** psachin has joined #openstack-swift | 06:51 | |
| *** mvalsecc has joined #openstack-swift | 07:15 | |
| *** mvalsecc has quit IRC | 07:15 | |
| *** rcernin has quit IRC | 08:09 | |
| *** rpittau|afk is now known as rpittau | 08:19 | |
| *** dsariel has joined #openstack-swift | 08:21 | |
| *** rcernin has joined #openstack-swift | 08:52 | |
| *** rcernin has quit IRC | 08:57 | |
| *** rcernin has joined #openstack-swift | 09:08 | |
| *** rcernin has quit IRC | 09:32 | |
| *** m75abrams has quit IRC | 09:42 | |
| *** dsariel has quit IRC | 09:54 | |
| *** rcernin has joined #openstack-swift | 09:57 | |
| *** rcernin has quit IRC | 10:02 | |
| *** m75abrams has joined #openstack-swift | 10:10 | |
| *** rcernin has joined #openstack-swift | 10:12 | |
| *** m75abrams has quit IRC | 10:14 | |
| *** m75abrams has joined #openstack-swift | 10:16 | |
| *** baojg has quit IRC | 10:16 | |
| *** rcernin has quit IRC | 10:22 | |
| *** dsariel has joined #openstack-swift | 10:42 | |
| zigo | After switching a container+account+proxy server to uwsgi, I got this in the container server logs: | 10:49 |
| zigo | Traceback (most recent call last): | 10:49 |
| zigo | File "/usr/lib/python3/dist-packages/eventlet/hubs/hub.py", line 461, in fire_timers | 10:49 |
| zigo | timer() | 10:49 |
| zigo | File "/usr/lib/python3/dist-packages/eventlet/hubs/timer.py", line 59, in __call__ | 10:49 |
| zigo | cb(*args, **kw) | 10:49 |
| zigo | File "/usr/lib/python3/dist-packages/eventlet/event.py", line 175, in _do_send | 10:49 |
| zigo | waiter.switch(result) | 10:49 |
| zigo | greenlet.error: cannot switch to a different thread | 10:49 |
| zigo | So I switched to a single thread. Though would there be a fix ? | 10:49 |
| zigo | It may kill performances no ? | 10:49 |
| *** rcernin has joined #openstack-swift | 11:01 | |
| *** rcernin has quit IRC | 11:01 | |
| *** rcernin has joined #openstack-swift | 11:04 | |
| *** baojg has joined #openstack-swift | 11:29 | |
| *** baojg has quit IRC | 11:29 | |
| *** baojg has joined #openstack-swift | 11:30 | |
| *** baojg has quit IRC | 11:30 | |
| *** baojg has joined #openstack-swift | 11:31 | |
| *** baojg has quit IRC | 11:31 | |
| *** baojg has joined #openstack-swift | 11:31 | |
| *** baojg has quit IRC | 11:32 | |
| *** baojg has joined #openstack-swift | 11:32 | |
| *** baojg has quit IRC | 11:33 | |
| *** baojg has joined #openstack-swift | 11:33 | |
| *** baojg has quit IRC | 11:33 | |
| *** rcernin has quit IRC | 11:58 | |
| *** rcernin has joined #openstack-swift | 12:06 | |
| *** dsariel has quit IRC | 12:17 | |
| *** rcernin has quit IRC | 12:22 | |
| *** psachin has quit IRC | 12:50 | |
| *** m75abrams has quit IRC | 12:55 | |
| *** dsariel has joined #openstack-swift | 13:09 | |
| *** m75abrams has joined #openstack-swift | 13:11 | |
| clayg | zigo: single os thread is probably correct - we normally use multiple async processes. N.B. https://uwsgi-docs.readthedocs.io/en/latest/Async.html does NOT call out eventlet as explicitly supported (but it's pretty similar in nature to gevent) | 13:32 |
| *** m75abrams has quit IRC | 13:49 | |
| acoles | zaitcev: thanks for questioning that. IIRC I needed a fix to make the bounds hashable, threw in that kludge and forgot to go back and think harder :( I think the problem actually stems from my failure to implement a proper singleton 3 years ago, so needing a __hash__ because of the __eq__ , and then forgetting that the classes are not singletons so a fixed hash is not appropriate. #willfix | 14:13 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: swift-manage-shard-ranges: add repair and analyze commands https://review.opendev.org/c/openstack/swift/+/765624 | 14:16 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Make ShardRange.OuterBound a proper singleton https://review.opendev.org/c/openstack/swift/+/782757 | 14:16 |
| acoles | zaitcev: so maybe that is better ^^^ | 14:16 |
| zaitcev | acoles: I thought something like md5(self.name) or whatever. Sorry, it was a drive-by, I didn't look well into it. | 14:59 |
| zaitcev | acoles: What I actually wanted to ask off you is where you used that class as a key in a dict. I didn't see it in the code. | 15:00 |
| acoles | it was a good call - I only needed __hash__ because we have an __eq__, and only need __eq__ because we failed to implement proper singleton | 15:01 |
| acoles | OIC, let me find a link | 15:01 |
| acoles | zaitcev: https://review.opendev.org/c/openstack/swift/+/765624/24/swift/container/sharder.py#361 | 15:04 |
| *** stand has joined #openstack-swift | 15:29 | |
| *** m75abrams has joined #openstack-swift | 15:34 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: relinker: Parallelize per disk https://review.opendev.org/c/openstack/swift/+/769633 | 16:00 |
| *** gyee has joined #openstack-swift | 16:23 | |
| *** timburke_ has joined #openstack-swift | 16:39 | |
| *** ChanServ sets mode: +v timburke_ | 16:39 | |
| *** timburke__ has quit IRC | 16:42 | |
| *** m75abrams has quit IRC | 17:02 | |
| *** rpittau is now known as rpittau|afk | 17:24 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: WIP: Enhance reclaim to estimate remaining tombstones https://review.opendev.org/c/openstack/swift/+/782832 | 17:51 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: WIP: sharder: update own shard range with tombstone estimate https://review.opendev.org/c/openstack/swift/+/782833 | 17:51 |
| *** rcernin has joined #openstack-swift | 18:19 | |
| *** rcernin has quit IRC | 18:23 | |
| *** timburke_ has quit IRC | 18:35 | |
| *** timburke__ has joined #openstack-swift | 18:35 | |
| *** ChanServ sets mode: +v timburke__ | 18:35 | |
| zaitcev | acoles: thanks a lot, I think I understand. We only have one instance of each MIN and MAX. | 19:24 |
| timburke__ | almost meeting time | 20:55 |
| seongsoocho | morning~~~ | 20:55 |
| kota_ | good morning | 20:58 |
| *** rcernin has joined #openstack-swift | 21:01 | |
| *** rcernin has quit IRC | 21:01 | |
| *** rcernin has joined #openstack-swift | 21:02 | |
| mattoliverau | morning | 21:03 |
| *** klamath_atx has quit IRC | 21:56 | |
| *** klamath_atx has joined #openstack-swift | 22:00 | |
| *** rcernin has quit IRC | 22:11 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: relinker: Add policy to relinker progress output https://review.opendev.org/c/openstack/swift/+/780440 | 22:14 |
| *** rcernin has joined #openstack-swift | 22:26 | |
| *** rcernin has quit IRC | 22:31 | |
| *** rcernin has joined #openstack-swift | 22:32 | |
| *** stand has quit IRC | 23:20 | |
| *** gyee has quit IRC | 23:35 | |
| *** Underknowledge has quit IRC | 23:40 | |
| *** Underknowledge has joined #openstack-swift | 23:51 | |
| *** Underknowledge has quit IRC | 23:54 | |
| *** Underknowledge has joined #openstack-swift | 23:54 | |
| *** Underknowledge has quit IRC | 23:59 | |
| *** Underknowledge has joined #openstack-swift | 23:59 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!