[O11y][SQL Input] Add healthcheck to all images #9861
Conversation
aliabbas-elastic
changed the title
|
/test |
| interval: 1s | ||
| retries: 90 |
There was a problem hiding this comment.
Running healthcheck every second (in case it is failing) is just too much unnecessary overhead.
Probably these are more sane defaults:
interval: 10s
timeout: <decide some sane default>
retries: 5
start_period: <decide some sane default>
If Postgres container on averages takes time to start up, let's set a start_period. If this healthcheck command using sqlcmd is quick, probably setting a small timeout is good. I think interval 5/10s and retries 5/10 should be more than enough.
There was a problem hiding this comment.
Yes. Setting interval and retries to values like 5/10 would make sense. But since the container does not take a lot time to start I think we can skip start_period and timeout IMO. Similar config was there in microsoft_sqlserver package but it's good to keep some proper values in retries and interval.
There was a problem hiding this comment.
I know. Many healthchecks are improper in our packages.
| USER root | ||
| ENV MYSQL_ROOT_PASSWORD test | ||
|
|
||
| HEALTHCHECK --interval=1s --retries=90 CMD /healthcheck.sh |
There was a problem hiding this comment.
Let's set better defaults. Retrying after every 1s is never good.
… into fix_sql_input_tests
|
/test |
💚 Build Succeeded
History
|
|
Proposed commit message
Fixes the flaky tests which result in error message
one or more errors found in documents stored in metrics-sql.sql-ep data stream: [0] found error.message in event: cannot open connection: testing connection: dial tcp 172.18.0.7:3306: connect: connection refusedChecklist
changelogas this change doesn't relate to the changes that should be visible at user's endHow to test this PR locally
Related issues
Screenshots