Hi,
Michael Paquier (running locally I think), and subsequently Thomas Munro
(noticing [1]), privately reported that they noticed an assertion failure in
GetSnapshotData(). Both reasonably were wondering if that's related to the
snapshot scalability patches.
Michael reported the following asserti
On Mon, May 31, 2021 at 09:37:17PM +0900, Michael Paquier wrote:
> I have been looking at all that for the last couple of days, and
> checked the code to make sure that relying on RecoveryInProgress() as
> the tipping point is logically correct in terms of virtual XID,
> snapshot build and KnownAss
On Fri, Oct 01, 2021 at 02:11:15PM +0900, Michael Paquier wrote:
> A couple of months later, I have looked back at this thread and this
> report. I have rechecked all the standby handling and snapshot builds
> involving KnownAssignedXids and looked at all the phases that are
> getting called until
On 2021-10-04 17:27:44 +0900, Michael Paquier wrote:
> On Fri, Oct 01, 2021 at 02:11:15PM +0900, Michael Paquier wrote:
> > A couple of months later, I have looked back at this thread and this
> > report. I have rechecked all the standby handling and snapshot builds
> > involving KnownAssignedXids
Hi Andres!
> 23 апр. 2021 г., в 01:36, Andres Freund написал(а):
>
> So snapshots within that window will always be "empty", i.e. xmin is
> latestCompletedXid and xmax is latestCompletedXid + 1. Once we reach 3), we'll
> look at the procarray, which then leads xmin going back to 588.
>
>
> I t
Hi,
On 2021-05-01 17:35:09 +0500, Andrey Borodin wrote:
> I'm investigating somewhat resemblant case.
> We have an OLTP sharded installation where shards are almost always under
> rebalancing. Data movement is implemented with 2pc.
> Switchover happens quite often due to datacenter drills. The in
> 3 мая 2021 г., в 23:10, Andres Freund написал(а):
>
> Hi,
>
> On 2021-05-01 17:35:09 +0500, Andrey Borodin wrote:
>> I'm investigating somewhat resemblant case.
>> We have an OLTP sharded installation where shards are almost always under
>> rebalancing. Data movement is implemented with 2p
Andres Freund writes:
> Michael Paquier (running locally I think), and subsequently Thomas Munro
> (noticing [1]), privately reported that they noticed an assertion failure in
> GetSnapshotData(). Both reasonably were wondering if that's related to the
> snapshot scalability patches.
> Michael rep
Hi,
On 2021-05-04 12:32:34 -0400, Tom Lane wrote:
> Andres Freund writes:
> > Michael Paquier (running locally I think), and subsequently Thomas Munro
> > (noticing [1]), privately reported that they noticed an assertion failure in
> > GetSnapshotData(). Both reasonably were wondering if that's r
On Thu, Apr 22, 2021 at 01:36:03PM -0700, Andres Freund wrote:
> The sequence in StartupXLOG() leading to the issue is the following:
>
> 1) RecoverPreparedTransactions();
> 2) ShutdownRecoveryTransactionEnvironment();
> 3) XLogCtl->SharedRecoveryState = RECOVERY_STATE_DONE;
>
> Because 2) resets
Hi,
On 2021-05-26 16:57:31 +0900, Michael Paquier wrote:
> Yes, there should not be any as far as I recall. 2PC is kind of
> special with its fake ProcArray entries.
It's really quite an awful design :(
> > I think to fix the issue we'd have to move
> > ShutdownRecoveryTransactionEnvironment()
On Thu, May 27, 2021 at 10:01:49AM -0700, Andres Freund wrote:
> Why would it be intrusive? We're talking a split second here, no? More
> importantly, I don't think it's correct to release the locks at that
> point.
I have been looking at all that for the last couple of days, and
checked the code
12 matches
Mail list logo