David Steele

> The reason query_id is important for auditing is because there are many cases where we can't guarantee the query text doesn't have sensitive data. That makes sense to me....

I understand that query_id could be useful but the fact that you need to build a dictionary separately to make it work (without also logging the query) is going to...

So this is ready for commit?

> It would be helpful to have this information in the pgbackrest info output. That seems useful but not sure how we'd go about implementing in. Lock files won't work...

It would be helpful to see the output from the `info` command. But at a guess setting `repo1-retention-archive=8` but `repo1-retention-full=4` is perhaps tricking pgBackRest into thinking it needs to keep...

As far as I can see you only have one full backup so the retention limit of four is not being met and archive expiration is being skipped. We do...

There is no logging to show why authentication is not correctly configured -- this is on purpose. But the `access denied` error is coming from pgbackrest so at least you...

> I see backup.info growing in size. (You said before that backup.info remain constant) Hmm, I don't think I said that. > This mean that the actual file contain all...

> The files are not re-generated and you end up with Yeah, the contributing guide was off here. Code generation for config happens when test.pl is run, not at compile...

This seems reasonable. Please submit a PR with the change and we'll have a look.