| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
| Cc: | pgsql-hackers(at)postgresql(dot)org, Paul van den Bogaard <Paul(dot)Vandenbogaard(at)Sun(dot)COM> |
| Subject: | Re: Reworking WAL locking |
| Date: | 2008-02-14 16:49:06 |
| Message-ID: | 9649.1203007746@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Simon Riggs <simon(at)2ndquadrant(dot)com> writes:
> LWLockAcquire(WALInsertLock, LW_EXCLUSIVE)
> Reserve space to write into.
> LSN = current Insert pointer
> Move pointer forward by length of data to be inserted, acquiring
> WALWriteLock if required to ensure space is available.
I think you've handwaved over how you'll avoid self-deadlock
in the case where the WAL record exceeds the size of wal_buffers.
Assuming that that can be fixed without too much ugliness,
it seems worth trying.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Guillaume Lelarge | 2008-02-14 18:27:46 | Re: Show INHERIT in \du |
| Previous Message | Tom Lane | 2008-02-14 16:42:17 | Re: Show INHERIT in \du |