| *** tkajinam is now known as Guest3966 | 00:09 | |
| *** tkajinam is now known as Guest3972 | 01:47 | |
| *** tkajinam is now known as Guest3986 | 05:43 | |
| *** tkajinam is now known as Guest3988 | 06:53 | |
| opendevreview | Alistair Coles proposed openstack/swift master: proxy: refactor error limiter to a class https://review.opendev.org/c/openstack/swift/+/858790 | 10:14 |
|---|---|---|
| opendevreview | Alistair Coles proposed openstack/swift master: Refactor memcache config and MemcacheRing loading https://review.opendev.org/c/openstack/swift/+/820648 | 10:14 |
| opendevreview | Alistair Coles proposed openstack/swift master: Global error limiter using memcache https://review.opendev.org/c/openstack/swift/+/820313 | 10:14 |
| edausq | hello, I have a question about auto_shard option of container-sharder. Configuration file and documentation say that it should not be used in production. Can you confirm it is still the case? | 14:20 |
| acoles | edausq: yes, that is still the recommendation. The reason is that sharding should only be initiated on a single replica of a container, but the auto_shard mode does not have a robust leader election process so is at risk of auto-initiating on more than one replica. | 21:02 |
| mattoliver | Yeah, don't turn on auto sharding yet. Using the sharing tool is fine (swift-manage-shard-ranges) | 21:03 |
| acoles | The recommended approach is to use the swift-manage-shard-ranges tool to initiate sharding on one replica (i.e. one DB). | 21:03 |
| mattoliver | Oh what Al said :) | 21:03 |
| mattoliver | I hope to get there one day, when I get some spare time to finally solve it. | 21:03 |
| mattoliver | I think know knowing the ring version will help (to rule out old primaries electing themselves), maybe having the ring history as well🤔 | 21:05 |
| acoles | edausq: swift-manage-shard-ranges can of course be called from an automated process outside of swift - the important thing is that something is acting as a single point of control for choosing one of the replica DBs to act on, and currently swift itself does not have any single point of control. | 21:05 |
| acoles | timburke__: is there a meeting today? | 21:06 |
| mattoliver | Good question! | 21:07 |
| timburke__ | oh, right! forgot. let's cancel for this week, do a ptg recap next week | 21:07 |
| timburke__ | (sorry, got distracted today on some home repairs) | 21:07 |
| kota | ok, back to my bed :P | 21:08 |
| mattoliver | Kk | 21:08 |
| timburke__ | thanks for getting up early kota and mattoliver! go to bed, acoles :-) | 21:08 |
| acoles | ok, it does feel like not much time has happened since the PTG :) | 21:08 |
| mattoliver | Enjoy the sleep in kota | 21:08 |
| kota | thx all | 21:09 |
| mattoliver | It's true, I've been distracted with mcrouter. Can't wait to get some swift time in. | 21:10 |
| edausq | acoles: mattoliver: thanks a lot for your feedbacks/comments | 21:30 |
| opendevreview | Tim Burke proposed openstack/swift master: CI: Install pip from pinned upstream for py2 probe tests https://review.opendev.org/c/openstack/swift/+/862758 | 22:15 |
| opendevreview | Tim Burke proposed openstack/swift stable/zed: Fix Zed CI https://review.opendev.org/c/openstack/swift/+/861629 | 22:37 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!