Re: Allow "snapshot too old" error, to prevent bloat

Amit Kapila <amit.kapila16@gmail.com> wrote:

> Could you please explain in slightly more detail why can't it work> if we use timestamp instead of snapshot->xmin in
yourpatch in
 

> function TestForOldSnapshot()?

It works fine for the additional visibility checking in scans, but
it doesn't cover the vacuuming -- that needs to use a transaction ID
for its cutoff.
> How exactly will this allow to return "snapshot old" error, do we
> need a check for page (page lsn) as in your current patch?


The change to the part where it actually throws the error is very
small, just checking time instead of xmin.


--
Kevin Grittner
EDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company