| From: | Takahiro Itagaki <itagaki(dot)takahiro(at)oss(dot)ntt(dot)co(dot)jp> |
|---|---|
| To: | Fujii Masao <masao(dot)fujii(at)gmail(dot)com> |
| Cc: | PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: lock mode for ControlFileLock which pg_start_backup uses |
| Date: | 2010-03-09 07:47:21 |
| Message-ID: | 20100309164721.99CA.52131E4D@oss.ntt.co.jp |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Fujii Masao <masao(dot)fujii(at)gmail(dot)com> wrote:
> Currently pg_start_backup() accesses the shared ControlFile
> by using ControlFileLock with LW_EXCLUSIVE lock mode. But
> since that access is read-only operation, LW_SHARED should
> be chosen instead of LW_EXCLUSIVE.
Almost all operations of ControlFileLock is in LW_EXCLUSIVE, but
there is one usage of LWLockConditionalAcquire(ControlFileLock, LW_SHARED)
in XLogNeedsFlush().
> The attached patch changes the lock mode which pg_start_backup()
> uses. Is it worth applying this patch?
I think the patch is reasonable to represent what we are doing,
even if there is no performance benefits from it.
Regards,
---
Takahiro Itagaki
NTT Open Source Software Center
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Takahiro Itagaki | 2010-03-09 07:55:09 | Re: Visual Studio 2005, C-language function - avoiding hacks? |
| Previous Message | Andrew Dunstan | 2010-03-09 02:52:57 | Re: invalid UTF-8 via pl/perl |