Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-35526

wsrep_sst_mariabackup script broken on 10.11.10

Details

    Description

      When upgrading one node in my Galera cluster from 10.11.9 to 10.11.10, i noticed that the upgraded node was unable to start correctly. The following lines were appearing in syslog"

      Nov 29 13:19:34 node02 -wsrep-sst-joiner: Cleanup after exit with status: 1
      Nov 29 13:19:34 node02 mariadbd[2661482]: 2024-11-29 13:19:34 0 [ERROR] WSREP: Failed to read 'ready <addr>' (got '(null)') from: wsrep_sst_mariabackup --role 'joiner' --address '10.97.149.12' --datadir '/var/lib/mysql/' --parent 2661482 --progress 0 --binlog '/var/log/mysql/mysql-bin' --binlog-index '/var/log/mysql/mariadb-bin.index' --mysqld-args --wsrep_start_position=e3a052f3-0858-11ef-b83e-67fa45d180b1:191
      Nov 29 13:19:34 node02 mariadbd[2661482]: 2024-11-29 13:19:34 0 [ERROR] WSREP: Process completed with error: wsrep_sst_mariabackup --role 'joiner' --address '10.97.149.12' --datadir '/var/lib/mysql/' --parent 2661482 --progress 0 --binlog '/var/log/mysql/mysql-bin' --binlog-index '/var/log/mysql/mariadb-bin.index' --mysqld-args --wsrep_start_position=e3a052f3-0858-11ef-b83e-67fa45d180b1:191: 1 (Operation not permitted)
      Nov 29 13:19:34 node02 mariadbd[2661482]: 2024-11-29 13:19:34 2 [ERROR] WSREP: Failed to prepare for 'mariabackup' SST. Unrecoverable.
      Nov 29 13:19:34 node02 mariadbd[2661482]: 2024-11-29 13:19:34 2 [ERROR] WSREP: SST request callback failed. This is unrecoverable, restart required.
      

      I've attached some more logging in an attachment

      I was able to fix replication by replacing the new script with the old script, used in 10.11.9, in /usr/bin/wsrep_sst_mariabackup

      Attachments

        Activity

          lveltmaat Hello, thanks for reporting this bug. Can you check the new scripts attached? I hope they fix this error. wsrep_sst_common.sh wsrep_sst_mariabackup.sh wsrep_sst_rsync.sh

          sysprg Julius Goryavsky added a comment - lveltmaat Hello, thanks for reporting this bug. Can you check the new scripts attached? I hope they fix this error. wsrep_sst_common.sh wsrep_sst_mariabackup.sh wsrep_sst_rsync.sh
          rtuk Rick Tuk added a comment -

          @Julius Goryavsky, I'm a colleague of @Lann Veltmaat, I've updated the scripts on one of our clusters running 10.11.10. Both a full transfer (removed all data from /var/lib/mysql) and an incremental transfer (reboot of a node) worked.
          I don't believe we use rsync in any way (please correct me if I'm wrong), so I'm not sure if the update of wsrep_sst_rsync.sh did anything / was included in the test, common and mariabackup are confirmed to be working in my opinion.
          Looking forward to the release, thank you for you quick action.

          rtuk Rick Tuk added a comment - @Julius Goryavsky, I'm a colleague of @Lann Veltmaat, I've updated the scripts on one of our clusters running 10.11.10. Both a full transfer (removed all data from /var/lib/mysql) and an incremental transfer (reboot of a node) worked. I don't believe we use rsync in any way (please correct me if I'm wrong), so I'm not sure if the update of wsrep_sst_rsync.sh did anything / was included in the test, common and mariabackup are confirmed to be working in my opinion. Looking forward to the release, thank you for you quick action.

          Fixed by the latest changes to the scripts that will be included in the 10.11.11 release.

          sysprg Julius Goryavsky added a comment - Fixed by the latest changes to the scripts that will be included in the 10.11.11 release.

          rtuk Thanks for testing, the rsync script is not needed in your case, but I attached it just in case (if you suddenly use this sst mode in some scenario), there are also some fixes compared to the previous release.

          sysprg Julius Goryavsky added a comment - rtuk Thanks for testing, the rsync script is not needed in your case, but I attached it just in case (if you suddenly use this sst mode in some scenario), there are also some fixes compared to the previous release.
          stephanvos Stephan Vos added a comment -

          Can you please indicate what in the script was broken.
          Will I experience the same issue when upgrading from 10.11.9 to 10.11.10?

          stephanvos Stephan Vos added a comment - Can you please indicate what in the script was broken. Will I experience the same issue when upgrading from 10.11.9 to 10.11.10?

          People

            sysprg Julius Goryavsky
            lveltmaat Lann Veltmaat
            Votes:
            0 Vote for this issue
            Watchers:
            6 Start watching this issue

            Dates

              Created:
              Updated:
              Resolved:

              Git Integration

                Error rendering 'com.xiplink.jira.git.jira_git_plugin:git-issue-webpanel'. Please contact your Jira administrators.