Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Thank you Robert for committing the patch. commit fc70a4b0df38bda6a13941f1581f25fbb643c7f3 I've changed the status to Committed. On Mon, Mar 27, 2017 at 6:09 AM, Michael Paquier wrote: > On Sat, Mar 25, 2017 at 5:26 PM, Kuntal Ghosh >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Sat, Mar 25, 2017 at 5:26 PM, Kuntal Ghosh wrote: > On Fri, Mar 24, 2017 at 9:23 PM, Robert Haas wrote: >> I think this is still not good. The places where pgstat_bestart() has >> been added are not even correct. For example, the call added

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Fri, Mar 24, 2017 at 9:23 PM, Robert Haas wrote: > On Thu, Mar 23, 2017 at 7:29 AM, Michael Paquier > wrote: >> On Thu, Mar 23, 2017 at 8:19 PM, Kuntal Ghosh >> wrote: >>> Hence, to be consistent with others,

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Thu, Mar 23, 2017 at 7:29 AM, Michael Paquier wrote: > On Thu, Mar 23, 2017 at 8:19 PM, Kuntal Ghosh > wrote: >> Hence, to be consistent with others, bgworker processes can be >> initialized from

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Thu, Mar 23, 2017 at 8:19 PM, Kuntal Ghosh wrote: > Hence, to be consistent with others, bgworker processes can be > initialized from BackgroundWorkerInitializeConnectionBy[Oid]. Yeah, I am fine with that. Thanks for the updated versions. -- Michael -- Sent via

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Wed, Mar 22, 2017 at 9:54 PM, Robert Haas wrote: > On Wed, Mar 22, 2017 at 12:20 PM, Robert Haas wrote: >> On Wed, Mar 22, 2017 at 1:31 AM, Michael Paquier >> wrote: >>> Okay, switched as ready for committer. One note

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Thu, Mar 23, 2017 at 1:24 AM, Robert Haas wrote: > I mean, your argument boils down to "somebody might want to > deliberately hide things from pg_stat_activity". But that's not > really a mode we support in general, and supporting it only for > certain cases doesn't

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Wed, Mar 22, 2017 at 12:20 PM, Robert Haas wrote: > On Wed, Mar 22, 2017 at 1:31 AM, Michael Paquier > wrote: >> Okay, switched as ready for committer. One note for the committer >> though: keeping the calls of pgstat_bestart() out of >>

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Wed, Mar 22, 2017 at 1:31 AM, Michael Paquier wrote: > Okay, switched as ready for committer. One note for the committer > though: keeping the calls of pgstat_bestart() out of > BackgroundWorkerInitializeConnection() and > BackgroundWorkerInitializeConnectionByOid()

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Mar 21, 2017 at 10:37 PM, Kuntal Ghosh wrote: > On Tue, Mar 21, 2017 at 10:52 AM, Michael Paquier > wrote: > We reserve a slot for each possible BackendId, plus one for each > possible auxiliary process type. For a non-auxiliary

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Mar 21, 2017 at 10:52 AM, Michael Paquier wrote: Thank you for the review. > Unfortunately this is true only for background workers that connect to > a database. And this would break for bgworkers that do not do that. > The point to fix is here: > + if

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Fri, Mar 17, 2017 at 6:19 PM, Kuntal Ghosh wrote: > On Thu, Mar 16, 2017 at 1:18 PM, Michael Paquier > wrote: >> + /* We have userid for client-backends and wal-sender processes */ >> + if (beentry->st_backendType == B_BACKEND || >>

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Thu, Mar 16, 2017 at 1:18 PM, Michael Paquier wrote: > On Wed, Mar 15, 2017 at 9:14 PM, Kuntal Ghosh > wrote: >> I've attached the updated patches. > > Thanks for the new versions. This begins to look really clear. Thanks again for the

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Wed, Mar 15, 2017 at 9:14 PM, Kuntal Ghosh wrote: > I've attached the updated patches. Thanks for the new versions. This begins to look really clear. > In 0001-Infra-to-expose-all-backend-processes-in-pg_stat_get.patch, > I've extended BackendStatusArray to to

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Mar 14, 2017 at 1:50 PM, Michael Paquier wrote: > "writer" would be better if defined as "background writer" instead? > You are forgetting in this list autovacuum workers and the startup > process, the latter is important for nodes in recovery. > Modified

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Wed, Mar 15, 2017 at 4:52 AM, Robert Haas wrote: > On Tue, Mar 14, 2017 at 7:22 AM, Kuntal Ghosh > wrote: >> I do have extended localBackendStatusTable with slots for non-backend >> processes. But, I've renamed it as localProcStatusTable

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Mar 14, 2017 at 7:22 AM, Kuntal Ghosh wrote: > I do have extended localBackendStatusTable with slots for non-backend > processes. But, I've renamed it as localProcStatusTable since it > includes all processes. I'll keep the variable name as >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Mar 14, 2017 at 1:50 PM, Michael Paquier wrote: > Here is a first pass on this patch. Thanks Michael for the review. > > void > -pgstat_bestart(void) > +pgstat_procstart(void) > I would not have thought that this patch justifies potentially > breaking

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Fri, Mar 10, 2017 at 2:36 PM, Kuntal Ghosh wrote: > On Fri, Mar 10, 2017 at 3:11 AM, Andres Freund wrote: >> On 2017-03-09 16:37:29 -0500, Tom Lane wrote: >>> Robert Haas writes: >>> > On Thu, Mar 9, 2017 at 2:30 PM,

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Fri, Mar 10, 2017 at 3:11 AM, Andres Freund wrote: > On 2017-03-09 16:37:29 -0500, Tom Lane wrote: >> Robert Haas writes: >> > On Thu, Mar 9, 2017 at 2:30 PM, Peter Eisentraut >> > wrote: >> >> In practice, I think

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On 2017-03-09 16:37:29 -0500, Tom Lane wrote: > Robert Haas writes: > > On Thu, Mar 9, 2017 at 2:30 PM, Peter Eisentraut > > wrote: > >> In practice, I think it's common to do a quick select * from > >> pg_stat_activity to determine

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Robert Haas writes: > On Thu, Mar 9, 2017 at 2:30 PM, Peter Eisentraut > wrote: >> In practice, I think it's common to do a quick select * from >> pg_stat_activity to determine whether a database instance is in use. > I thought of the

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Hi, On 2017-03-09 14:30:21 -0500, Peter Eisentraut wrote: > In practice, I think it's common to do a quick select * from > pg_stat_activity to determine whether a database instance is in use. > (You always see your own session, but that's easy to eyeball.) If we > add all the various background

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Thu, Mar 9, 2017 at 2:30 PM, Peter Eisentraut wrote: > Perhaps I'm confused by the title of this thread/CF entry, but > background workers already do show up in pg_stat_activity. (You can > verify that by testing the background sessions patch.) So which >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Perhaps I'm confused by the title of this thread/CF entry, but background workers already do show up in pg_stat_activity. (You can verify that by testing the background sessions patch.) So which additional things are we aiming to see with this? In practice, I think it's common to do a quick

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Hello Amit, On Tue, Mar 7, 2017 at 4:24 PM, Amit Langote wrote: > Hi Kuntal, > > Patches apply and compile fine. Works as advertised. > > Some minor comments on the patches themselves. > Thanks for the review. > In 0001: > > - * pgstat_bestart() - > + *

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Hi Kuntal, Patches apply and compile fine. Works as advertised. Some minor comments on the patches themselves. In 0001: - * pgstat_bestart() - + * pgstat_procstart() - + * + * Initialize this process's entry in the PgBackendStatus array. + * Called from InitPostgres and

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Hello everyone, As discussed in this thread, I've attached a set of patches to show auxiliary processes, autovacuum launcher and bgworker along with other backends in pg_stat_activity. For now, I've extended BackendStatusArray to store auxiliary processes. Backends use slots indexed in the range

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Dec 13, 2016 at 11:40:54AM -0500, Robert Haas wrote: > Let's confine ourselves to fixing one problem at a time. I think we > can get where we want to be in this case by adding one new column and > some new rows to pg_stat_activity. Agreed. Let's also remove the abuse of WAL senders with

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Mon, Dec 12, 2016 at 6:45 PM, Kevin Grittner wrote: > On Mon, Dec 12, 2016 at 10:33 AM, Andres Freund > wrote: > > On 2016-12-12 13:26:32 -0300, Alvaro Herrera wrote: > >> Robert Haas wrote: > > >>> 1. Show all processes that have a PGPROC in

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Mon, Dec 12, 2016 at 8:13 PM, Michael Paquier wrote: > On Tue, Dec 13, 2016 at 10:05 AM, Craig Ringer wrote: >> We should probably expose a proc_type or something, with types: >> >> * client_backend >> * bgworker >> * walsender >> * autovacuum

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On 13 December 2016 at 09:13, Michael Paquier wrote: > On Tue, Dec 13, 2016 at 10:05 AM, Craig Ringer wrote: >> We should probably expose a proc_type or something, with types: >> >> * client_backend >> * bgworker >> * walsender >> * autovacuum >>

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Dec 13, 2016 at 10:05 AM, Craig Ringer wrote: > We should probably expose a proc_type or something, with types: > > * client_backend > * bgworker > * walsender > * autovacuum > * checkpointer > * bgwriter A text field is adapted then, more than a single character.

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On 13 December 2016 at 01:45, Kevin Grittner wrote: > On Mon, Dec 12, 2016 at 10:33 AM, Andres Freund wrote: >> On 2016-12-12 13:26:32 -0300, Alvaro Herrera wrote: >>> Robert Haas wrote: > 1. Show all processes that have a PGPROC in pg_stat_activity, >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Dec 13, 2016 at 1:45 AM, Robert Haas wrote: > And now I'm noticing that Michael Paquier previously started a thread > on this problem which I failed to note before starting this one: > >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Mon, Dec 12, 2016 at 10:33 AM, Andres Freund wrote: > On 2016-12-12 13:26:32 -0300, Alvaro Herrera wrote: >> Robert Haas wrote: >>> 1. Show all processes that have a PGPROC in pg_stat_activity, >>> 2. Add a second view, say pg_stat_system_activity, >> I vote 1. > > +1

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Mon, Dec 12, 2016 at 11:19 AM, Robert Haas wrote: > So, one of the problems in this patch as committed is that for any > process that doesn't show up in pg_stat_activity, there's no way to > see the wait event information. That sucks. I think there are > basically two

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Mon, Dec 12, 2016 at 9:19 AM, Robert Haas wrote: > So, one of the problems in this patch as committed is that for any > process that doesn't show up in pg_stat_activity, there's no way to > see the wait event information. That sucks. I think there are > basically two

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On 2016-12-12 13:26:32 -0300, Alvaro Herrera wrote: > Robert Haas wrote: > > > So, one of the problems in this patch as committed is that for any > > process that doesn't show up in pg_stat_activity, there's no way to > > see the wait event information. That sucks. I think there are > >

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Alvaro Herrera writes: > Robert Haas wrote: >> So, one of the problems in this patch as committed is that for any >> process that doesn't show up in pg_stat_activity, there's no way to >> see the wait event information. That sucks. I think there are >> basically two

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

Robert Haas wrote: > So, one of the problems in this patch as committed is that for any > process that doesn't show up in pg_stat_activity, there's no way to > see the wait event information. That sucks. I think there are > basically two ways to fix this: > > 1. Show all processes that have a

Re: [HACKERS] exposing wait events for non-backends (was: Tracking wait event for latches)

On Tue, Oct 4, 2016 at 11:59 AM, Robert Haas wrote: > On Mon, Oct 3, 2016 at 8:43 PM, Michael Paquier > wrote: >> The rest looks good to me. Thanks for the feedback and the time! > > Thanks for the fixes. I committed this with an additional