Investigate the network IO and latency of executing sql query and store results |
||
Issue descriptionIf network IO is too large or latency is too high for each run, we may want to implement resume point in sql query to filter out previous detected flakes.
,
Jan 8
I tried to implement such a resume point, but it turns out to be complicate, and I remember the main issue is that it's hard to find a field that can be used to filter out processed entries in the table. Given that we're moving to expose and retrieve failed tests from the recipe side, which is a much cleaner solution, I don't think it's worth doing this anymore. |
||
►
Sign in to add a comment |
||
Comment 1 by chanli@chromium.org
, Dec 27