Uploaded image for project: 'MariaDB Server'
  1. MariaDB Server
  2. MDEV-18271

Galera 4: test manually rolling upgrade to Server 10.4 + Galera 4

Details

    Description

      Steps:

      1. Set up Galera cluster using branch 10.4 or 10.3 with 3 nodes.
      2. Shutdown one node.
      3. Upgrade the binaries used by that node to bb-10.4-galera4 and galera 4 library.
      4. Start that server.

      Expected: Upgraded node has joined the cluster.

      5. Repeat the same for the rest of nodes.

      Attachments

        Issue Links

          relates to

          Bug - A problem which impairs or prevents the functions of the product. MDEV-18407 Galera: Rolling upgrade: 10.3 nodes are stopped with signal 6 on attempt to join upgraded 10.4 node

          • Blocker - Blocks development and/or testing work, production could not run.
          • Closed

          Task - A task that needs to be done. MDEV-18504 Galera 4: test manually bulk upgrade to Server 10.4 + Galera 4

          • Critical - Crashes, loss of data, severe memory leak.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. MDEV-18580 Galera: Rolling upgrade: Upgraded node 3 is stopped with signal 6 after node 2 shutdown

          • Critical - Crashes, loss of data, severe memory leak.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. MDEV-18629 Galera: Rolling upgrade: Upgraded node is stopped with signal 6 on attempt to join the cluster

          • Blocker - Blocks development and/or testing work, production could not run.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. MDEV-18699 Galera: Rolling upgrade: Upgraded node is stopped on commit if wsrep_trx_fragment_size > 0

          • Critical - Crashes, loss of data, severe memory leak.
          • Closed

          Bug - A problem which impairs or prevents the functions of the product. MDEV-19983 Galera: Rolling upgrade: Upgraded node 2 cannot connect to the cluster on rows adding, updating, and removing during upgrade

          • Critical - Crashes, loss of data, severe memory leak.
          • Closed

          Activity

            Ascending order - Click to sort in descending order
            stepan.patryshev Stepan Patryshev (Inactive) added a comment -

            Status as of 27.02.2019.

            The following scenarios have been tested:

            1. PASSED: Without data loading during upgrade.
            2. PASSED: Data loading during upgrade.
            3. PASSED: Data updating and removing during upgrade.
            4. FAILED: Schema changing during upgrade. Added a table and a column. Should be documented as a known limitation.
            5. FAILED: Streaming replication on the partially upgraded cluster. See MDEV-18699 and MDEV-18552 for it.
            6. Streaming replication on upgraded cluster. It worked for me, but shako has found an issue MDEV-18631 using sysbench.

            stepan.patryshev Stepan Patryshev (Inactive) added a comment - Status as of 27.02.2019. The following scenarios have been tested: 1. PASSED: Without data loading during upgrade. 2. PASSED: Data loading during upgrade. 3. PASSED: Data updating and removing during upgrade. 4. FAILED: Schema changing during upgrade. Added a table and a column. Should be documented as a known limitation. 5. FAILED: Streaming replication on the partially upgraded cluster. See MDEV-18699 and MDEV-18552 for it. 6. Streaming replication on upgraded cluster. It worked for me, but shako has found an issue MDEV-18631 using sysbench.
            stepan.patryshev Stepan Patryshev (Inactive) added a comment -

            Current status: MDEV-19983: Galera: Rolling upgrade: Upgraded node 2 cannot connect to the cluster on rows adding, updating, and removing during upgrade.

            stepan.patryshev Stepan Patryshev (Inactive) added a comment - Current status: MDEV-19983 : Galera: Rolling upgrade: Upgraded node 2 cannot connect to the cluster on rows adding, updating, and removing during upgrade.
            stepan.patryshev Stepan Patryshev (Inactive) added a comment -

            Rolling upgrade from CS 10.3 to 10.4 has been performed successfully on the following versions:
            All binaries are non-debug built from sources:

            MariaDB Server 10.3.19-MariaDB-debug: branch 10.3, commit 90a9c4cae74d2ef1008e3f216026b7fd2697e46b.
            Galera Lib3 25.3.28(r3875): branch mariadb-3.x, commit 92d6990b65259d12327c211e6658cf1c0c818cfb.

            MariaDB Server 10.4.9-MariaDB-debug: branch 10.4, commit b9dea911bf8e3d4d8fc57ce3ef15ab0e2ab4d076.
            Galera Lib4 26.4.3(r4535): branch mariadb-4.x, commit 752664dc3c7065d8e0c73ac99d0028a5f84eb250.

            stepan.patryshev Stepan Patryshev (Inactive) added a comment - Rolling upgrade from CS 10.3 to 10.4 has been performed successfully on the following versions: All binaries are non-debug built from sources: MariaDB Server 10.3.19-MariaDB-debug: branch 10.3, commit 90a9c4cae74d2ef1008e3f216026b7fd2697e46b. Galera Lib3 25.3.28(r3875): branch mariadb-3.x, commit 92d6990b65259d12327c211e6658cf1c0c818cfb. MariaDB Server 10.4.9-MariaDB-debug: branch 10.4, commit b9dea911bf8e3d4d8fc57ce3ef15ab0e2ab4d076. Galera Lib4 26.4.3(r4535): branch mariadb-4.x, commit 752664dc3c7065d8e0c73ac99d0028a5f84eb250.
            stepan.patryshev Stepan Patryshev (Inactive) added a comment - - edited

            Closing as done.

            stepan.patryshev Stepan Patryshev (Inactive) added a comment - - edited Closing as done.

            People

              stepan.patryshev Stepan Patryshev (Inactive)
              stepan.patryshev Stepan Patryshev (Inactive)
              Votes:
              0 Vote for this issue
              Watchers:
              3 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.