Thread

  1. PgBackRest client_loop: send disconnect: Connection reset

    KK CHN <kkchn.in@gmail.com> — 2024-08-29T13:31:48Z

    List,
    
    I am facing an Error :  on a connection reset while  PgBackRest is doing
    backup process for DB server to Repo Server.
    
    But the reissue of the     pgbackrest backup  command  fails as follows...
    
    "Unable to acquire lock on file '/tmp/pgbackrest/Repo-backup.lock':
    Resource temporarily unavailable        HINT: is another pgBackRest process
    running?
    
    
    After multiple attempts and  during  8 to 10 minutes,   the backup command
    from RepoServer starts working again..    How to avoid /what is the exact
    reason for   "Resource temporarily unavailable"?  HINT : Is another
    PgBackRest process running ?
    
    
    Please advise ..
    
    On Connection reset then issuing backup command I get out puts as below.
    
    
    ######################################################################
    
    [root@dbtest ~]# sudo -u postgres pgbackrest --stanza=Repo
    --log-level-console=info backup
    2024-08-29 16:38:13.922 P00   INFO: backup command begin 2.52.1: --delta
    --exec-id=658924-527c843b --log-level-console=info --log-level-file=debug
    --pg1-host=10.15.0.202 --pg1-host-user=enterprisedb
    --pg1-path=/data/edb/as16/data --pg-version-force=16 --process-max=5
    --repo1-block --repo1-bundle --repo1-cipher-pass=<redacted>
    --repo1-cipher-type=aes-256-cbc --repo1-path=/data/DB_BKUPS
    --repo1-retention-diff=2 --repo1-retention-full=2 --stanza=Repo --start-fast
    WARN: no prior backup exists, incr backup has been changed to full
    2024-08-29 16:38:16.745 P00   INFO: execute non-exclusive backup start:
    backup begins after the requested immediate checkpoint completes
    2024-08-29 16:38:18.003 P00   INFO: backup start archive =
    00000001000000850000009A, lsn = 85/9A0000D8
    2024-08-29 16:38:18.003 P00   INFO: check archive for prior segment
    000000010000008500000099
    WARN: resumable backup 20240829-105625F of same type exists -- invalid
    files will be removed then the backup will resume
    client_loop: send disconnect: Connection reset
    
    
    [root@dbtest ~]# sudo -u postgres pgbackrest --stanza=Repo
    --log-level-console=info backup
    2024-08-29 18:45:06.630 P00   INFO: backup command begin 2.52.1: --delta
    --exec-id=691718-6c37ff5c --log-level-console=info --log-level-file=debug
    --pg1-host=10.15.0.202 --pg1-host-user=enterprisedb
    --pg1-path=/data/edb/as16/data --pg-version-force=16 --process-max=5
    --repo1-block --repo1-bundle --repo1-cipher-pass=<redacted>
    --repo1-cipher-type=aes-256-cbc --repo1-path=/data/DB_BKUPS
    --repo1-retention-diff=2 --repo1-retention-full=2 --stanza=Repo --start-fast
    ERROR: [050]: unable to acquire lock on file
    '/tmp/pgbackrest/Repo-backup.lock': Resource temporarily unavailable
           HINT: is another pgBackRest process running?
    
    2024-08-29 18:45:06.631 P00   INFO: backup command end: aborted with
    exception [050]
    [root@dbtest ~]# sudo -u postgres pgbackrest --stanza=Repo
    --log-level-console=info backup
    2024-08-29 18:45:14.531 P00   INFO: backup command begin 2.52.1: --delta
    --exec-id=691730-06e11989 --log-level-console=info --log-level-file=debug
    --pg1-host=10.15.0.202 --pg1-host-user=enterprisedb
    --pg1-path=/data/edb/as16/data --pg-version-force=16 --process-max=5
    --repo1-block --repo1-bundle --repo1-cipher-pass=<redacted>
    --repo1-cipher-type=aes-256-cbc --repo1-path=/data/DB_BKUPS
    --repo1-retention-diff=2 --repo1-retention-full=2 --stanza=Repo --start-fast
    ERROR: [050]: unable to acquire lock on file
    '/tmp/pgbackrest/Repo-backup.lock': Resource temporarily unavailable
           HINT: is another pgBackRest process running?
    2024-08-29 18:45:14.531 P00   INFO: backup command end: aborted with
    exception [050]
    
    ###########################################################################
    After 10 minutes or so I issued the command again as follows it starts
    backing up as follows  ?
    
    
    
    
    2024-08-29 18:48:19.561 P00   INFO: backup command end: aborted with
    exception [050]
    [root@dbtest ~]# sudo -u postgres pgbackrest --stanza=Repo
    --log-level-console=info backup
    2024-08-29 18:52:33.834 P00   INFO: backup command begin 2.52.1: --delta
    --exec-id=693517-4994fba1 --log-level-console=info --log-level-file=debug
    --pg1-host=10.15.0.202 --pg1-host-user=enterprisedb
    --pg1-path=/data/edb/as16/data --pg-version-force=16 --process-max=5
    --repo1-block --repo1-bundle --repo1-cipher-pass=<redacted>
    --repo1-cipher-type=aes-256-cbc --repo1-path=/data/DB_BKUPS_NGERSS
    --repo1-retention-diff=2 --repo1-retention-full=2 --stanza=Repo --start-fast
    WARN: no prior backup exists, incr backup has been changed to full
    2024-08-29 18:52:36.813 P00   INFO: execute non-exclusive backup start:
    backup begins after the requested immediate checkpoint completes
    2024-08-29 18:52:37.788 P00   INFO: backup start archive =
    000000010000008600000063, lsn = 86/63000028
    2024-08-29 18:52:37.788 P00   INFO: check archive for prior segment
    000000010000008600000062
    WARN: resumable backup 20240829-105625F of same type exists -- invalid
    files will be removed then the backup will resume
    
    
    Thank you,
    Krishane
    
  2. Re: PgBackRest client_loop: send disconnect: Connection reset

    Greg Sabino Mullane <htamfids@gmail.com> — 2024-08-29T14:11:26Z

    On Thu, Aug 29, 2024 at 9:31 AM KK CHN <kkchn.in@gmail.com> wrote:
    
    > "Unable to acquire lock on file '/tmp/pgbackrest/Repo-backup.lock'
    >
    
    When this happens, take a look inside this file. If there is another
    pgbackrest process running, the pid will be inside that file. Kill that
    process before trying to run another backup.
    
    Cheers,
    Greg
    
  3. Re: PgBackRest client_loop: send disconnect: Connection reset

    Ron <ronljohnsonjr@gmail.com> — 2024-08-29T14:47:23Z

    On Thu, Aug 29, 2024 at 10:12 AM Greg Sabino Mullane <htamfids@gmail.com>
    wrote:
    
    > On Thu, Aug 29, 2024 at 9:31 AM KK CHN <kkchn.in@gmail.com> wrote:
    >
    >> "Unable to acquire lock on file '/tmp/pgbackrest/Repo-backup.lock'
    >>
    >
    > When this happens, take a look inside this file. If there is another
    > pgbackrest process running, the pid will be inside that file. Kill that
    > process before trying to run another backup.
    >
    
    Or don't run pgbackrest when another instance is running.  This is what I
    do:
    pgrep -a pgbackrest | grep -q "pgbackrest backup" && { echo "another
    pgbackrest back is running"; exit 1; }
    
    (There's probably a better way to check that, but this works.)
    
    -- 
    Death to America, and butter sauce.
    Iraq lobster!