mirror of
https://github.com/nspcc-dev/neofs-node.git
synced 2026-03-01 04:29:10 +00:00
s3 PUT errors on AZ when we rebooted VEDI. #885
Labels
No labels
I1
I2
I3
I4
S0
S1
S2
S3
S4
U0
U1
U2
U3
U4
blocked
bug
config
dependencies
discussion
documentation
enhancement
enhancement
epic
feature
go
good first issue
help wanted
neofs-adm
neofs-cli
neofs-cli
neofs-cli
neofs-ir
neofs-lens
neofs-storage
neofs-storage
performance
question
security
task
test
windows
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
nspcc-dev/neofs-node#885
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @anikeev-yadro on GitHub (Nov 14, 2022).
Originally assigned to: @anikeev-yadro on GitHub.
Expected Behavior
We expect that reboot of one node not affected s3 load on other nodes.
Steps to Reproduce (for bugs)
1.Create s3 buckets and load objects with policy "node-off": "REP 2 CBF 2 SELECT 2 FROM *"
2.Start k6 R/W load on node AZ
3.Reboot node VEDI
4.Have a lot of PUT errors in k6:
As result PUT fails on 18 objects, which were not putted.
Logs and preload files: TEMP.zip
Version:
Your Environment
Server setup and configuration:
cloud, 4 VMs, 4 SN, 4 http qw, 4 s3 gw
Operating System and version (uname -a):
linux vedi 5.10.0-16-amd64 https://github.com/nspcc-dev/neofs-node/issues/1 SMP Debian 5.10.127-1 (2022-06-30) x86_64 GNU/Linux
@fyrchik commented on GitHub (Nov 15, 2022):
For the first 3 requests:
context cancellederrors are expected under load in some configurations.@fyrchik commented on GitHub (Nov 18, 2022):
S3 timeout issue
@anikeev-yadro commented on GitHub (Nov 21, 2022):
This is the not timeout issue, beacuse change
stream_timeoutnot addressed issue.stream_timeout=10s and object size=64Kb, 18 operations was failed
stream_timeout=20s, obect seize=64Kb, one operation was failed:
but when we increased obejct size to 10Mb (stream_timeout=20s) we had 29 fails again:
@fyrchik commented on GitHub (Nov 21, 2022):
@anikeev-yadro What was the
stream_timeoutin bothneofs-storageands3-gw?@anikeev-yadro commented on GitHub (Nov 25, 2022):
neofs-storage:
s3-gw:
@anikeev-yadro commented on GitHub (Dec 30, 2022):
Not reproduced on 10 VUs in k6, but reproduced always on >40 VUs and 201 Mb object size.
@fyrchik commented on GitHub (Dec 31, 2022):
It could always be reproduced on some number of writers.