Re: "bad key in cancel request" - Mailing list pgsql-hackers

From Tom Lane
Subject Re: "bad key in cancel request"
Date
Msg-id 14028.1194192362@sss.pgh.pa.us
Whole thread Raw
In response to "bad key in cancel request"  (Neil Conway <neilc@samurai.com>)
Responses Re: "bad key in cancel request"
List pgsql-hackers
Neil Conway <neilc@samurai.com> writes:
>     ereport(DEBUG2,
>             (errmsg_internal("bad key in cancel request for process %d",
>              backendPID)));

> I think this ought to be logged at a higher level than DEBUG2: for one
> thing, it is a potential security issue the DBA might want to be aware
> of. It could also indicate a misconfigured or buggy client application.
> Therefore, I think we should raise the level to WARNING, or perhaps
> NOTICE. Any objections?

No, if it's intended for the log it should be LOG.  Your other proposals
are actually *less* likely to get to where the DBA could see them.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Greg Smith
Date:
Subject: Re: Test lab
Next
From: Zdenek Kotala
Date:
Subject: Re: Intel x64 vs AMD x64 pgdata