| gmann | timburke: i think i pinged for this bug which you already commented in bug i will debug from tempest side and let you know if need more info https://bugs.launchpad.net/swift/+bug/1905432 | 00:08 |
|---|---|---|
| openstack | Launchpad bug 1905432 in tempest "test_create_object_with_transfer_encoding failing since 20th Nov for " [Undecided,New] | 00:08 |
| timburke | 👍 | 00:11 |
| *** jv has joined #openstack-swift | 01:01 | |
| *** timburke_ has joined #openstack-swift | 01:27 | |
| *** ChanServ sets mode: +v timburke_ | 01:27 | |
| *** timburke has quit IRC | 01:28 | |
| *** timburke_ has quit IRC | 01:32 | |
| *** rcernin has quit IRC | 03:04 | |
| *** rcernin has joined #openstack-swift | 03:18 | |
| *** rcernin has quit IRC | 03:21 | |
| *** rcernin has joined #openstack-swift | 03:22 | |
| *** psachin has joined #openstack-swift | 03:35 | |
| seongsoocho | Does anyone know about disk performance tuning? I tested disk io performance in object node, the result were incomprehensible.. | 03:40 |
| seongsoocho | In /srv/node/sdd, the disk read bandwidth is about 2.7 GB/s. But In object directory, the read performance drops sharply.. | 03:40 |
| seongsoocho | this is my test | 03:40 |
| seongsoocho | [root@sto sdd] # dd if=/srv/node/sdd/test.file of=/dev/zero bs=64M | 03:41 |
| seongsoocho | 100+0 records in | 03:41 |
| seongsoocho | 100+0 records out | 03:41 |
| seongsoocho | 6710886400 bytes (6.7 GB) copied, 2.47789 s, 2.7 GB/s | 03:41 |
| seongsoocho | [root@sto d7bc20111175964820c5752c18e6cc13]# dd if=/srv/node/sdc/objects/3451/c13/d7bc20111175964820c5752c18e6cc13/1606753037.78666.data of=/dev/null bs=64M | 03:41 |
| seongsoocho | 58+1 records in | 03:41 |
| seongsoocho | 58+1 records out | 03:41 |
| seongsoocho | 3937929216 bytes (3.9 GB) copied, 22.0082 s, 179 MB/s | 03:41 |
| seongsoocho | oops, I missed. "dd if=/srv/node/sdc/test.file of=/dev/zero bs=64M" is also same result | 03:42 |
| *** timburke has joined #openstack-swift | 03:55 | |
| *** ChanServ sets mode: +v timburke | 03:55 | |
| *** timburke has quit IRC | 03:59 | |
| *** manuvakery_ has joined #openstack-swift | 05:05 | |
| *** rcernin has quit IRC | 05:22 | |
| *** rcernin has joined #openstack-swift | 05:32 | |
| *** evrardjp has quit IRC | 05:33 | |
| *** evrardjp has joined #openstack-swift | 05:33 | |
| openstackgerrit | Michele Valsecchi proposed openstack/swift master: Fail if swift binary are missing https://review.opendev.org/c/openstack/swift/+/764137 | 05:49 |
| *** gyee has quit IRC | 05:55 | |
| *** rcernin has quit IRC | 06:12 | |
| *** m75abrams has joined #openstack-swift | 06:28 | |
| *** klamath_atx has quit IRC | 06:47 | |
| *** timburke has joined #openstack-swift | 06:53 | |
| *** ChanServ sets mode: +v timburke | 06:53 | |
| *** timburke has quit IRC | 07:53 | |
| *** rpittau|afk is now known as rpittau | 07:59 | |
| *** takamatsu has quit IRC | 08:13 | |
| *** takamatsu has joined #openstack-swift | 08:15 | |
| *** takamatsu has quit IRC | 08:53 | |
| *** takamatsu has joined #openstack-swift | 08:54 | |
| *** psachin has quit IRC | 09:07 | |
| *** mvalsecc has quit IRC | 09:08 | |
| *** adeberg has quit IRC | 09:09 | |
| *** dsariel has quit IRC | 09:11 | |
| *** dsariel has joined #openstack-swift | 09:12 | |
| *** adeberg has joined #openstack-swift | 09:12 | |
| *** dsariel has quit IRC | 09:32 | |
| *** dsariel has joined #openstack-swift | 09:32 | |
| *** jv has quit IRC | 10:17 | |
| *** manuvakery_ has quit IRC | 12:22 | |
| *** baojg has joined #openstack-swift | 13:16 | |
| *** jv has joined #openstack-swift | 13:20 | |
| *** baojg has quit IRC | 13:30 | |
| *** baojg has joined #openstack-swift | 13:31 | |
| *** m75abrams has quit IRC | 14:55 | |
| *** m75abrams has joined #openstack-swift | 14:57 | |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Only merge shard ranges from root when SHRINKING or SHRUNK https://review.opendev.org/c/openstack/swift/+/764163 | 15:16 |
| openstackgerrit | Alistair Coles proposed openstack/swift master: Add probe test coverage for shard overlap resolution https://review.opendev.org/c/openstack/swift/+/764951 | 15:16 |
| *** timburke has joined #openstack-swift | 15:43 | |
| *** ChanServ sets mode: +v timburke | 15:43 | |
| *** timburke has quit IRC | 15:46 | |
| *** timburke has joined #openstack-swift | 15:46 | |
| *** ChanServ sets mode: +v timburke | 15:46 | |
| *** irclogbot_2 has quit IRC | 15:53 | |
| *** irclogbot_3 has joined #openstack-swift | 15:57 | |
| *** timburke_ has joined #openstack-swift | 16:08 | |
| *** ChanServ sets mode: +v timburke_ | 16:08 | |
| *** timburke has quit IRC | 16:11 | |
| *** irclogbot_3 has quit IRC | 16:11 | |
| *** irclogbot_2 has joined #openstack-swift | 16:13 | |
| *** m75abrams has quit IRC | 16:47 | |
| *** rpittau is now known as rpittau|afk | 17:01 | |
| clayg | seongsoocho: some kind of file system fragmentation? if we could keep track of the age of a device in the ring we could periodically drain/format/readd to help performance and let aging sectors out before corruption sets in | 17:10 |
| *** gyee has joined #openstack-swift | 17:23 | |
| *** takamatsu has quit IRC | 19:19 | |
| *** klamath_atx has joined #openstack-swift | 19:54 | |
| klamath_atx | Howdy, I was wondering if there is any issue using tempauth and keystone auth on the same swift cluster? Running into some odd response codes when using curl vs swift cli. Curl is returning 200OK while swift-cli is returning 403 using the same endpoint and username/password combos | 19:56 |
| *** rcernin has joined #openstack-swift | 21:24 | |
| *** Diplomat has joined #openstack-swift | 21:47 | |
| Diplomat | Hey guys. I'm following Victora's guide to install Swift and I'm working on building those rings, but I'm getting a fancy big python error.. https://dpaste.org/0D3J is that normal ? | 21:48 |
| Diplomat | Looks like there's a similar bug https://bugs.launchpad.net/tripleo/+bug/1836572 | 21:57 |
| openstack | Launchpad bug 1836572 in tripleo "[swift][python3] InterpolationSyntaxError: '%' must be followed by '%' or '(', found: '%SWIFT_HASH_PATH_SUFFIX%'" [High,Fix released] - Assigned to yatin (yatinkarel) | 21:57 |
| Diplomat | Yea, documentation didn't have a notice to update the swift.conf.. I set the hash there and ran "swift-ring-builder account.builder create 10 3 1" and it worked fine again | 22:01 |
| *** gyee has quit IRC | 22:04 | |
| *** gyee has joined #openstack-swift | 22:05 | |
| rledisez | seongsoocho: except if you're running on SSD/NVMe, 179 MB/s (1.4 Gb/s) is what I would expect from a HDD on linear access. I guess your first test was reading from VFS cache instead of the HDD. you may want to flush cache first, run this before retesting: echo 3 > /proc/sys/vm/drop_caches | 22:17 |
| rledisez | If you want more representative test, look for fio or something similar | 22:17 |
| timburke_ | klamath_atx, you should be able to run with both auth systems in the same cluster, yes. we've got gate jobs that do it all the time. need to make sure auth_token has delay_auth_decision enabled, but you'd need that to make other things (like tempurl) work, anyway. order might matter; looks like tests have a pipeline like `... s3api authtoken s3token keystoneauth tempauth ...` | 22:20 |
| timburke_ | if it's still not working, let us know! important things to know will be proxy-server config, whether it's tempauth or keystone users having trouble (or both), and how swift-cli is getting its auth info (maybe tryit with `--debug` and `--retries 0` if the error is pretty reliable?) | 22:24 |
| DHE | I have an older version of swift running with both. no problems. just gotta remember the delay_auth_decision (?) option | 22:39 |
| Diplomat | Hey guys. Does anyone happen to have any idea why I'm getting such error: "Service Unavailable (HTTP 503) (Request-ID: tx2b7606069b6e455eac020-005fc6c6cb)" <-- I get this from the controller when I'm trying to upload a text file. There's nothing in the object server (no logs in swift dir) but I can see PUT requests in messages log file , but no errors. | 22:43 |
| Diplomat | This transaction id matches 3 PUT requests, but that's about it.. no errors. Other actions like container creation or container listing work just fine | 22:44 |
| Diplomat | Ops, I had container ring set up wrong.. wrong port. It's all good now | 23:25 |
| *** Diplomat has left #openstack-swift | 23:26 | |
| *** aluria has quit IRC | 23:33 | |
| *** aluria has joined #openstack-swift | 23:34 | |
| *** klamath_atx has quit IRC | 23:39 | |
| openstackgerrit | Tim Burke proposed openstack/swift master: Clean up requirements a bit https://review.opendev.org/c/openstack/swift/+/764856 | 23:43 |
| *** klamath_atx has joined #openstack-swift | 23:47 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!