(dynamic) e.getMessage() from SSTableAcquisitionException

Error message

(dynamic) e.getMessage() from SSTableAcquisitionException

What it means

During PendingAntiCompaction's acquireTuple, Cassandra tries to mark candidate sstables compacting for a repair session. If it fails with SSTableAcquisitionException (e.g. sstables already compacting or being acquired elsewhere), the message is logged at WARN while the full stack trace is only at DEBUG, and the acquisition attempt returns null so repair proceeds without those sstables this round.

Source

Thrown at src/java/org/apache/cassandra/db/repair/PendingAntiCompaction.java:210

        private AcquireResult acquireTuple()
        {
            // this method runs with compactions stopped & disabled
            try
            {
                // using predicate might throw if there are conflicting ranges
                Set<SSTableReader> sstables = cfs.getLiveSSTables().stream().filter(predicate).collect(Collectors.toSet());
                if (sstables.isEmpty())
                    return new AcquireResult(cfs, null, null);

                LifecycleTransaction txn = cfs.getTracker().tryModify(sstables, OperationType.ANTICOMPACTION);
                if (txn != null)
                    return new AcquireResult(cfs, Refs.ref(sstables), txn);
                else
                    logger.error("Could not mark compacting for {} (sstables = {}, compacting = {})", sessionID, sstables, cfs.getTracker().getCompacting());
            }
            catch (SSTableAcquisitionException e)
            {
                logger.warn(e.getMessage());
                logger.debug("Got exception trying to acquire sstables", e);
            }

            return null;
        }

        protected AcquireResult acquireSSTables()
        {
            return cfs.runWithCompactionsDisabled(this::acquireTuple, predicate, OperationType.ANTICOMPACTION, false, false, false);
        }

        public AcquireResult call()
        {
            logger.debug("acquiring sstables for pending anti compaction on session {}", sessionID);
            // try to modify after cancelling running compactions. This will attempt to cancel in flight compactions including the given sstables for
            // up to a minute, after which point, null will be returned
            long start = currentTimeMillis();
            long delay = TimeUnit.SECONDS.toMillis(acquireRetrySeconds);

View on GitHub (pinned to 88fd0f6a0e)

Solutions

Check debug.log for the full 'Got exception trying to acquire sstables' stack to see which sstables/compactions collided
Avoid running multiple repairs concurrently; serialize repair sessions
Throttle compaction (nodetool stop compaction temporarily or compaction_throughput) during repair windows
Retry the repair after compactions quiesce; the message itself indicates the node handled it gracefully

Defensive patterns

Strategy: retry

Validate before calling

// avoid collisions: check compacting set before repair
Set<SSTableReader> compacting = cfs.getTracker().getCompacting();
if (compacting.containsAll(candidates)) logger.warn("Candidates busy; defer repair");

Try / catch

// caller pattern around PendingAntiCompaction
try { result = antiCompactionTask.get(); } catch (SSTableAcquisitionException e) { logger.info("Retry later: {}", e.getMessage()); }

Prevention

Serialize repair sessions; avoid overlapping repairs
Quiesce compaction during repair windows
Track pending compactions before scheduling repairs
Check debug.log for which sstables collided

When it happens

Trigger: Running incremental repair (nodetool repair) while concurrent compactions or another repair session already holds the required sstables via the CompactionStrategyManager lifecycle transaction.

Common situations: Overlapping repairs on the same table; aggressive concurrent compaction throughput; anticompaction racing with foreground compaction on busy nodes.

Understand the failure class

Background: Conflicting config options: "cannot be used together" — configuration validation errors across open-source libraries — this error's family across 162 libraries.

Related errors

AI-assisted analysis of apache/cassandra@88fd0f6a0e (2026-09-10). Data as JSON: /api/errors/3f1d4bacdfef00c1. Report an issue: GitHub.