Standard internal error counter for all exporters?

[RichiH]

I think it would make sense to have a canonical metric for internal error counting of all exporters. This would allow easy monitoring when scrape happens upon a partially misconfigured or otherwise non-ideal target.

This is partially motivated by prometheus/snmp_exporter#16

[brian-brazil]

There's so many ways that an exporter can fail that this isn't practical. For example that issue is a user configuration error, which is different from the target having an error which is different from a failure.

At most we can offer only general guidance.

[RichiH]

While I readily agree that there is a myriad of reasons, the point is that this would allow me to look at anything that is not zero.

[RichiH]

#1169 is somewhat related to this.

[brian-brazil]

#1169 is to cover a failure of the exporter itself to handle a failure.

[RichiH]

Correct. One is for failures that could be handled/ignored, one is for the ones which could not.

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.