Re: NOLOGGING option, or ?

Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> Recent test results have shown a substantial performance improvement
> (+25%) if WAL logging is disabled for large COPY statements.

How much of that is left after we fix the 64-bit-CRC issue?

> Now, I would like to discuss adding an enable_logging USERSET GUC,

[ fear and loathing ... ]

I don't like the idea of a GUC at all, and USERSET is right out.
I think it would have to be system-wide (cf fsync) to be even
implementable let alone somewhat predictable. Even if it could
be done per-backend with reasonable semantics, random users should
not get to make that decision --- it should be the DBA's call,
which means it needs at least SUSET permissions.

BTW, I'm sure you are the last one who needs to be reminded that
any such thing breaks PITR completely. Which is surely sufficient
reason not to let it be USERSET.

regards, tom lane