| From: | Simon Riggs <simon(at)2ndQuadrant(dot)com> | 
|---|---|
| To: | Fujii Masao <masao(dot)fujii(at)gmail(dot)com> | 
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> | 
| Subject: | Re: WAL Restore process during recovery | 
| Date: | 2012-01-17 10:31:47 | 
| Message-ID: | CA+U5nMJBVWPJAk0=V46pffp0OQ668x53vBhCcOa+pPre3kekzg@mail.gmail.com | 
| Views: | Whole Thread | Raw Message | Download mbox | Resend email | 
| Thread: | |
| Lists: | pgsql-hackers | 
On Tue, Jan 17, 2012 at 6:52 AM, Fujii Masao <masao(dot)fujii(at)gmail(dot)com> wrote:
> On Mon, Jan 16, 2012 at 2:06 AM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> WALRestore process asynchronously executes restore_command while
>> recovery continues working.
>>
>> Overlaps downloading of next WAL file to reduce time delays in file
>> based archive recovery.
>>
>> Handles cases of file-only and streaming/file correctly.
>
> Though I've not reviewed the patch deeply yet, I observed the following
> two problems when I tested the patch.
>
> When I set up streaming replication + archive (i.e., restore_command is set)
> and started the standby, I got the following error:
>
>    FATAL:  all AuxiliaryProcs are in use
>    LOG:  walrestore process (PID 18839) exited with exit code 1
>
> When I started an archive recovery without setting restore_command,
> it successfully finished.
Oh, I did have NUM_AUXILIARY_PROCS increased at one point, but I
"realised" it wasn't needed and removed it. Will change that. Thanks.
-- 
 Simon Riggs                   http://www.2ndQuadrant.com/
 PostgreSQL Development, 24x7 Support, Training & Services
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Noah Misch | 2012-01-17 10:33:22 | Re: Collect frequency statistics for arrays | 
| Previous Message | Heikki Linnakangas | 2012-01-17 10:16:34 | Re: BGWriter latch, power saving |