Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-11-01 Thread Jeff
On Oct 29, 2012, at 12:42 PM, Jeff Janes wrote: > On Mon, Oct 29, 2012 at 6:05 AM, Albe Laurenz wrote: >> I am configuring streaming replication with hot standby >> with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). >> PostgreSQL was compiled from source. >> >> It works fine, excep

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-11-01 Thread Jeff Trout
On Oct 29, 2012, at 12:42 PM, Jeff Janes wrote: > On Mon, Oct 29, 2012 at 6:05 AM, Albe Laurenz wrote: >> I am configuring streaming replication with hot standby >> with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). >> PostgreSQL was compiled from source. >> >> It works fine, excep

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread k...@rice.edu
On Tue, Oct 30, 2012 at 02:16:57PM +0100, Albe Laurenz wrote: > k...@rice.edu wrote: > >>> If you do not have good random io performance log replay is nearly > >>> unbearable. > >>> > >>> also, what io scheduler are you using? if it is cfq change that to > >>> deadline or noop. > >>> that can make

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread Albe Laurenz
k...@rice.edu wrote: >>> If you do not have good random io performance log replay is nearly >>> unbearable. >>> >>> also, what io scheduler are you using? if it is cfq change that to >>> deadline or noop. >>> that can make a huge difference. >> >> We use the noop scheduler. >> As I said, an identic

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread Albe Laurenz
Heikki Linnakangas wrote: >> Why does WAL replay read much more than it writes? >> I thought that pretty much every block read during WAL >> replay would also get dirtied and hence written out. > > Not necessarily. If a block is modified and written out of the buffer > cache before next checkpoint

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread k...@rice.edu
On Tue, Oct 30, 2012 at 09:50:44AM +0100, Albe Laurenz wrote: > >> On Mon, Oct 29, 2012 at 6:05 AM, Albe Laurenz > wrote: > >>> I am configuring streaming replication with hot standby > >>> with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). > >>> PostgreSQL was compiled from source. >

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread Heikki Linnakangas
On 30.10.2012 10:50, Albe Laurenz wrote: Why does WAL replay read much more than it writes? I thought that pretty much every block read during WAL replay would also get dirtied and hence written out. Not necessarily. If a block is modified and written out of the buffer cache before next checkp

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-30 Thread Albe Laurenz
>> On Mon, Oct 29, 2012 at 6:05 AM, Albe Laurenz wrote: >>> I am configuring streaming replication with hot standby >>> with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). >>> PostgreSQL was compiled from source. >>> >>> It works fine, except that starting the standby took for ever: >>

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-29 Thread Jeff Janes
On Mon, Oct 29, 2012 at 6:05 AM, Albe Laurenz wrote: > I am configuring streaming replication with hot standby > with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). > PostgreSQL was compiled from source. > > It works fine, except that starting the standby took for ever: > it took the s

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-29 Thread Albe Laurenz
Alvaro Herrera wrote: >> I am configuring streaming replication with hot standby >> with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). >> PostgreSQL was compiled from source. >> >> It works fine, except that starting the standby took for ever: >> it took the system more than 80 minutes

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-29 Thread Alvaro Herrera
Albe Laurenz wrote: > I am configuring streaming replication with hot standby > with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). > PostgreSQL was compiled from source. > > It works fine, except that starting the standby took for ever: > it took the system more than 80 minutes to rep

Re: [PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-29 Thread k...@rice.edu
On Mon, Oct 29, 2012 at 02:05:24PM +0100, Albe Laurenz wrote: > I am configuring streaming replication with hot standby > with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). > PostgreSQL was compiled from source. > > It works fine, except that starting the standby took for ever: > it t

[PERFORM] Replaying 48 WAL files takes 80 minutes

2012-10-29 Thread Albe Laurenz
I am configuring streaming replication with hot standby with PostgreSQL 9.1.3 on RHEL 6 (kernel 2.6.32-220.el6.x86_64). PostgreSQL was compiled from source. It works fine, except that starting the standby took for ever: it took the system more than 80 minutes to replay 48 WAL files and connect to