Add job ID to error table #2245
Conversation
Jamesbarford
force-pushed
the
fix/error-table
branch
from
86652c5 to
f6acb98
Compare
…`job_id` and rename columns Pass through an `Option<u32>` for job id to record errors fix postgres-to-sqlite fix insert statements for job_id more fixes for postgres-to-sqlite fixes for sqlite-to-postgres sqlite does not support `SERIAL` add an id column for the error converter add an job_id column for the error converter
Jamesbarford
force-pushed
the
fix/error-table
branch
from
f6acb98 to
148f177
Compare
collector/src/bin/collector.rs
Outdated
| // not with a benchmark. | ||
| conn.record_error( | ||
| artifact_row_id, | ||
| &format!("job:{}", benchmark_job.id()), |
There was a problem hiding this comment.
We can change the context now, because the job ID will be stored explicitly in the table now. Not sure what the context should be named in this case though, maybe just "unknown" or "general".
There was a problem hiding this comment.
Let's use something more general please as the context, retry count exhaustion is not the only thing that can happen, even now we have four different kinds of permanent errors. Maybe just Job failure, to distinguish it from benchmark failures.
…le and remove `IF NOT EXISTS`
I'm not sure which docs file you are referring too? |
|
|
Jamesbarford
force-pushed
the
fix/error-table
branch
from
7801679 to
f2ece09
Compare
I thought I'd updated it? I've changed |
|
Your updates are fine - I wrote that message before you did the MD docs update 😅 Maybe you just some some outdated GitHub comment? I'm fine with the current state of the PR :) Tried it locally and it seems to work fine. |
2 participants
job_idto the error table, remove unique constraint.benchmarkanderrortocontextandmessage.job_idin collector code.