| opendevreview | Jianjian Huo proposed openstack/swift master: common: add memcached based cooperative token mechanism. https://review.opendev.org/c/openstack/swift/+/890174 | 04:34 |
|---|---|---|
| opendevreview | Jianjian Huo proposed openstack/swift master: proxy: use cooperative tokens to coalesce updating shard range requests into backend https://review.opendev.org/c/openstack/swift/+/908969 | 05:29 |
| opendevreview | Alistair Coles proposed openstack/swift master: statsd: deprecate log_ prefix for options https://review.opendev.org/c/openstack/swift/+/922518 | 11:42 |
| opendevreview | Alistair Coles proposed openstack/swift master: remove unused log_headers option from docs https://review.opendev.org/c/openstack/swift/+/925762 | 11:42 |
| opendevreview | Alistair Coles proposed openstack/swift master: expirer: fix missing unit test assertions https://review.opendev.org/c/openstack/swift/+/925482 | 13:01 |
| opendevreview | Alistair Coles proposed openstack/swift master: sq? simplify Fake[Labeled]StatsdCLient https://review.opendev.org/c/openstack/swift/+/925780 | 15:10 |
| opendevreview | Alistair Coles proposed openstack/swift master: object-server: return 503 not 404 if meta or data file unlinked https://review.opendev.org/c/openstack/swift/+/909674 | 15:35 |
| timburke | DHE, that does *sound* like a race -- are you sure the PUT got back a 200 before the GET 404ed? have you checked how many frags were successfully written? you should be able to read your writes as long as it was to a new object name... or was the PUT an overwrite (so the GET should either return old or new data, but *not* 404)? | 16:25 |
| DHE | we generally don't overwrite anything. I'm either using the swift CLI (though an old version), or waiting for curl to return. I figure either way I've gotten a 201, right? | 16:36 |
| DHE | and I mean CLI curl command | 16:36 |
| timburke | strange -- yeah, i'd expect you to be able to read the write as long as it was after the upload completed. maybe the PUT went to deeper handoffs than expected because of error-limiting, then the GET had a different set of error-limited nodes? as long as even *one* durable frag was found on GET, though, it should have 503ed... | 17:18 |
| DHE | The cluster is fairly healthy... a few drives lately having issues as we cross that magic moment in time when their warranty expires and somehow the spinning metal knows it.... | 17:39 |
| DHE | 300 disks total, give or take a bit | 17:39 |
| opendevreview | Tim Burke proposed openstack/swift master: Remove legacy bin/ scripts https://review.opendev.org/c/openstack/swift/+/924795 | 18:00 |
| opendevreview | Merged openstack/swift master: expirer stats: add test coverage for 'errors' stat https://review.opendev.org/c/openstack/swift/+/925575 | 18:25 |
| opendevreview | Tim Burke proposed openstack/swift master: Add helpers for adding/popping parameters https://review.opendev.org/c/openstack/swift/+/925825 | 22:55 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!