apache/druid · error · SegmentLoadingException

Couldn't delete segments from Azure. See the task logs for…

Error message

Couldn't delete segments from Azure. See the task logs for more details.

What it means

AzureDataSegmentKiller.kill aggregates per-batch segment deletions; if any batch or individual delete failed (batchSuccessful false or an exception was caught), it throws this generic SegmentLoadingException pointing the operator at task logs. The actual per-segment failure reason is logged, not carried in this exception.

Solutions

Inspect the task/service logs for the per-segment BlobStorageException logged before this throw
Re-run the kill; failures are often transient or caused by already-deleted segments
Verify the Azure credentials have delete (blob write) permissions on the container
Check no concurrent kill tasks operate on the same segments

Example fix

// before
if (shouldThrowException) {
  throw new SegmentLoadingException("Couldn't delete segments from Azure. See the task logs for more details.");
}
// after
if (shouldThrowException) {
  throw new SegmentLoadingException(null, "Couldn't delete segments from Azure; last error: %s", lastErrorMessage);
}

Defensive patterns

Strategy: try-catch

Validate before calling

// verify permissions up front
boolean canDelete = testDeletePermission(container, prefix); // list + attempt small delete with a probe blob

Try / catch

try { killer.kill(segment); } catch (SegmentLoadingException e) { log.error("Azure kill failed, see cause logs: %s", e.getMessage()); /* retry later or alert */ }

Prevention

Grant delete permissions on the container to the Druid storage account
Avoid concurrent kill tasks on the same segments
Enable idempotent retry of kills; deletes are safe to re-run
Check task logs immediately — this exception deliberately omits the root cause

When it happens

Trigger: kill() or killBatch() fails to delete one or more segment blobs: emptyCloudBlobDirectory or deleteBlob throws (BlobStorageException, missing blob directory, permissions), or a batch reports partial failure.

Common situations: Killing segments whose files were already deleted (concurrent kill), storage account key lacking delete permission, or transient Azure 412/409 conflicts during batch kills.

Understand the failure class

Background: "API error: {status}" and "HTTP 401/403/404/429/5xx" errors: non-2xx HTTP responses explained — this error's family across 27 libraries.

Related errors

AI-assisted analysis of apache/druid@9b90983fd2 (2026-09-07). Data as JSON: /api/errors/24cb7884b23cb334. Report an issue: GitHub.

Appendix: source

Thrown at extensions-core/azure-extensions/src/main/java/org/apache/druid/storage/azure/AzureDataSegmentKiller.java:107

      );
      keysToDelete.add(blobPath);
    }

    boolean shouldThrowException = false;
    for (Map.Entry<String, List<String>> containerToKeys : containerToKeysToDelete.entrySet()) {
      boolean batchSuccessful = azureStorage.batchDeleteFiles(
              containerToKeys.getKey(),
              containerToKeys.getValue(),
              null
      );

      if (!batchSuccessful) {
        shouldThrowException = true;
      }
    }

    if (shouldThrowException) {
      throw new SegmentLoadingException(
          "Couldn't delete segments from Azure. See the task logs for more details."
      );
    }
  }


  @Override
  public void kill(DataSegment segment) throws SegmentLoadingException
  {
    log.info("Killing segment [%s]", segment);

    Map<String, Object> loadSpec = segment.getLoadSpec();
    final String containerName = MapUtils.getString(loadSpec, "containerName");
    final String blobPath = MapUtils.getString(loadSpec, "blobPath");
    final String dirPath = Paths.get(blobPath).getParent().toString();

    try {
      azureStorage.emptyCloudBlobDirectory(containerName, dirPath);

View on GitHub (pinned to 9b90983fd2)