Cassandra
  1. Cassandra
  2. CASSANDRA-3629

Bootstrapping nodes don't ensure schema is ready before continuing

    Details

    • Type: Bug Bug
    • Status: Resolved
    • Priority: Major Major
    • Resolution: Fixed
    • Fix Version/s: 1.0.7
    • Component/s: Core
    • Labels:
      None

      Description

      A bootstrapping node will assume that after it has slept for RING_DELAY it has all of the schema migrations and can continue the bootstrap process. However, with a large enough amount of migrations this is not sufficient and causes problems.

        Activity

        Hide
        Jonathan Ellis added a comment -

        +1

        should we put this in 0.8.9 as well?

        Show
        Jonathan Ellis added a comment - +1 should we put this in 0.8.9 as well?
        Hide
        Brandon Williams added a comment -

        should we put this in 0.8.9 as well?

        The 0.8.9 ship looks sailed unless we want to reroll for this, do you mean 0.8.10?

        Committed to 1.0, the 0.8 backport isn't trivial in any case, so leaving the ticket open for that.

        Show
        Brandon Williams added a comment - should we put this in 0.8.9 as well? The 0.8.9 ship looks sailed unless we want to reroll for this, do you mean 0.8.10? Committed to 1.0, the 0.8 backport isn't trivial in any case, so leaving the ticket open for that.
        Hide
        Jonathan Ellis added a comment -

        If it's a big backport, let's leave it as a known limitation of 0.8. We need to keep 0.8 stable more than we need this fixed there.

        Show
        Jonathan Ellis added a comment - If it's a big backport, let's leave it as a known limitation of 0.8. We need to keep 0.8 stable more than we need this fixed there.
        Hide
        Jonathan Ellis added a comment -

        The easiest workaround for this is CASSANDRA-3600, which is also 1.0-only. That much should be safe to backport...

        Show
        Jonathan Ellis added a comment - The easiest workaround for this is CASSANDRA-3600 , which is also 1.0-only. That much should be safe to backport...
        Hide
        Joaquin Casares added a comment - - edited

        A valid workaround is to save your current schema, drain and stop all the nodes, move all Migration and Schema sstables from all nodes to another directory (for temporary backup), stop all nodes, restart one seed node, paste the old schema (to create the first mutation and recreate the schema), restart that node one more time, then start the rest of the nodes.

        Show
        Joaquin Casares added a comment - - edited A valid workaround is to save your current schema, drain and stop all the nodes, move all Migration and Schema sstables from all nodes to another directory (for temporary backup), stop all nodes, restart one seed node, paste the old schema (to create the first mutation and recreate the schema), restart that node one more time, then start the rest of the nodes.

          People

          • Assignee:
            Brandon Williams
            Reporter:
            Brandon Williams
            Reviewer:
            Jonathan Ellis
          • Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

            • Created:
              Updated:
              Resolved:

              Development