Обсуждение: pg_basebackup state?
I need help determining the state of the pg_basebackup command. This works command works on other hosts and I make a connection as recorded in the postmaster log: 2015-07-02 12:10:53 EDT,0,authentication LOG: replication connection authorized: user=rep the process is running: > ps -ef | grep pg_base pg 32600 11558 0 12:10 pts/1 00:00:00 pg_basebackup -D basebackup/ --xlog --format=t -P -U rep pg 32601 1473 0 12:10 ? 00:00:00 postgres: wal sender process rep [local] sending backup "pg_basebackup base backup" but no data is written. Looks like wal sender is hurt. max_wal_senders = 3, I wonder if this is the reason for the stall. How to eval if this is set to low? Thanks.
On 7/2/15 12:39 PM, Ray Stell wrote: > I need help determining the state of the pg_basebackup command. This > works command works on other hosts and I make a connection as recorded > in the postmaster log: > > 2015-07-02 12:10:53 EDT,0,authentication LOG: replication connection > authorized: user=rep > > the process is running: > > ps -ef | grep pg_base > pg 32600 11558 0 12:10 pts/1 00:00:00 pg_basebackup -D > basebackup/ --xlog --format=t -P -U rep > pg 32601 1473 0 12:10 ? 00:00:00 postgres: wal sender > process rep [local] sending backup "pg_basebackup base backup" > > but no data is written. Looks like wal sender is hurt. > max_wal_senders = 3, I wonder if this is the reason for the stall. How > to eval if this is set to low? Thanks. v 9.4.4, I meant to include.
On 07/02/2015 09:39 AM, Ray Stell wrote: > > I need help determining the state of the pg_basebackup command. This > works command works on other hosts and I make a connection as recorded > in the postmaster log: > > 2015-07-02 12:10:53 EDT,0,authentication LOG: replication connection > authorized: user=rep > > the process is running: > > ps -ef | grep pg_base > pg 32600 11558 0 12:10 pts/1 00:00:00 pg_basebackup -D > basebackup/ --xlog --format=t -P -U rep > pg 32601 1473 0 12:10 ? 00:00:00 postgres: wal sender > process rep [local] sending backup "pg_basebackup base backup" > > but no data is written. Looks like wal sender is hurt. max_wal_senders > = 3, I wonder if this is the reason for the stall. How to eval if this > is set to low? Thanks. It may be that pg_basebackup is waiting for a checkpoint to complete. Try adding -cfast. JD -- Command Prompt, Inc. - http://www.commandprompt.com/ 503-667-4564 PostgreSQL Centered full stack support, consulting and development. Announcing "I'm offended" is basically telling the world you can't control your own emotions, so everyone else should do it for you.
Le 2 juil. 2015 6:41 PM, "Ray Stell" <stellr@vt.edu> a écrit :
>
> I need help determining the state of the pg_basebackup command. This works command works on other hosts and I make a connection as recorded in the postmaster log:
>
> 2015-07-02 12:10:53 EDT,0,authentication LOG: replication connection authorized: user=rep
>
> the process is running:
> > ps -ef | grep pg_base
> pg 32600 11558 0 12:10 pts/1 00:00:00 pg_basebackup -D basebackup/ --xlog --format=t -P -U rep
> pg 32601 1473 0 12:10 ? 00:00:00 postgres: wal sender process rep [local] sending backup "pg_basebackup base backup"
>
> but no data is written. Looks like wal sender is hurt. max_wal_senders = 3, I wonder if this is the reason for the stall. How to eval if this is set to low? Thanks.
>
You wouldn't have the wal sender if it was too low.
On 7/2/15 12:50 PM, Joshua D. Drake wrote: > > On 07/02/2015 09:39 AM, Ray Stell wrote: >> >> I need help determining the state of the pg_basebackup command. This >> works command works on other hosts and I make a connection as recorded >> in the postmaster log: >> >> 2015-07-02 12:10:53 EDT,0,authentication LOG: replication connection >> authorized: user=rep >> >> the process is running: >> > ps -ef | grep pg_base >> pg 32600 11558 0 12:10 pts/1 00:00:00 pg_basebackup -D >> basebackup/ --xlog --format=t -P -U rep >> pg 32601 1473 0 12:10 ? 00:00:00 postgres: wal sender >> process rep [local] sending backup "pg_basebackup base backup" >> >> but no data is written. Looks like wal sender is hurt. max_wal_senders >> = 3, I wonder if this is the reason for the stall. How to eval if this >> is set to low? Thanks. > > It may be that pg_basebackup is waiting for a checkpoint to complete. > Try adding -cfast. > > JD yep, that got it. thanks.