session_test: Increase tracing info timeout #966
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
While at it, we should decide what to do about #709 |
|
|
Lorak-mmk
force-pushed
the
fix-tracing-defaults
branch
2 times, most recently
from
aaeef26 to
963daa7
Compare
Added comments about Cassandra in session builder, and increased retries amount from 5 to 10, I think we can close #709 with this. |
wprzytula
approved these changes
Mar 19, 2024
Lorak-mmk
force-pushed
the
fix-tracing-defaults
branch
from
963daa7 to
d0e8ca2
Compare
Default values for fetching tracing info are quite strict in the driver: it will make 5 attempts to fetch the data, waiting 3ms between attempts. For Scylla this is enough, but for Cassandra it is not. For this reason our tests sometimes fail. This was partially (for one test) fixed previously: scylladb@61ad878 This commit fixes rest of the tests in session_test that could be affected. It also changes timeout parameters. Previous fixed used 50 attempts with 200ms wait, I changed it to 200 attempts with 50ms wait in order to not slow down Scylla tests unnecessarily.
Cassandra sometimes takes a lot of time to update tracing tables. Default values in session builder are suitable for Scylla, so Cassandra users may encounter unexpected timeouts. This commit adds a note suggesting them to increase the values.
Default values are a bit optimistic: fetching will be attempted 5 times, with 3ms sleeps between attempts. I am reluctant to increase sleep time, because it will increase latency in optimistic case. This commit increases default attempts number from 5 to 10, so that fetching tracing is more reliable, bot not overly lengthy.
Lorak-mmk
force-pushed
the
fix-tracing-defaults
branch
from
d0e8ca2 to
c70faec
Compare
wprzytula
approved these changes
Mar 19, 2024
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Default values for fetching tracing info are quite strict in the driver: it will make 5 attempts to fetch the data, waiting 3ms between attempts.
For Scylla this is enough, but for Cassandra it is not. For this reason our tests sometimes fail.
This was partially (for one test) fixed previously: 61ad878
This PR fixes other test case that might be affected.
Previous fix used 50 attempts with 200ms wait, I changed it to 200 attempts with 50ms wait in order to not slow down Scylla tests unnecessarily.
Increased default retry attempts in Session from 5 to 10 to make tracing fetching more reliable for Scylla users without increasing latency.
I didn't touch default sleep between attempts - for Scylla new values should be ok, and Cassandra users can change them in config. I added a warning in docs so that they are aware of the problem.
Fixes: #957
Fixes: #709
Pre-review checklist
./docs/source/.Fixes:annotations to PR description.