eolivelli commented on a change in pull request #1902: (WIP) Metadata checker
validating EnsemblePlacementpolicy
URL: https://github.com/apache/bookkeeper/pull/1902#discussion_r248784328
##########
File path:
bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/Auditor.java
##########
@@ -803,8 +885,72 @@ void checkAllLedgers() throws BKException, IOException,
InterruptedException, Ke
LOG.error("Got exception while trying to set
checkAllLedgersCTime", ue);
}
} finally {
- admin.close();
- client.close();
+ localAdmin.close();
+ localClient.close();
+ }
+ }
+
+ void metadataCheck() throws BKAuditException {
+ final CountDownLatch metadataCheckLatch = new CountDownLatch(1);
+ Processor<Long> ledgerProcessor = new Processor<Long>() {
+ @Override
+ public void process(Long ledgerId, AsyncCallback.VoidCallback
iterCallback) {
+
ledgerManager.readLedgerMetadata(ledgerId).whenComplete((metadataVer,
exception) -> {
+ if (exception == null) {
+ LedgerMetadata metadata = metadataVer.getValue();
+ int writeQuorumSize = metadata.getWriteQuorumSize();
+ int ackQuorumSize = metadata.getAckQuorumSize();
+ if (metadata.isClosed()) {
+ for (Map.Entry<Long, ? extends
List<BookieSocketAddress>> ensemble : metadata
+ .getAllEnsembles().entrySet()) {
+ long startEntryIdOfSegment = ensemble.getKey();
+ List<BookieSocketAddress> ensembleOfSegment =
ensemble.getValue();
+ boolean segmentAdheringToPlacementPolicy =
admin.isEnsembleAdheringToPlacementPolicy(
+ ensembleOfSegment, writeQuorumSize,
ackQuorumSize);
+ if (!segmentAdheringToPlacementPolicy) {
+
metadataCheckEnsembleNotAdheringToPlacementPolicy.inc();
+ LOG.warn(
+ "For ledger: {}, Segment starting
at entry: {}, with ensemble: {} having "
+ + "writeQuorumSize: {} and
ackQuorumSize: {} is not adhering to "
+ +
"EnsemblePlacementPolicy",
+ ledgerId, startEntryIdOfSegment,
ensembleOfSegment, writeQuorumSize,
+ ackQuorumSize);
+ }
+ }
+ } else {
+ LOG.debug("Ledger: {} is not yet closed, so
skipping the metadata check analysis for now",
+ ledgerId);
+ }
+ iterCallback.processResult(BKException.Code.OK, null,
null);
+ } else if (BKException
+ .getExceptionCode(exception) ==
BKException.Code.NoSuchLedgerExistsException) {
+ LOG.debug("Ignoring replication of already deleted
ledger {}", ledgerId);
+ iterCallback.processResult(BKException.Code.OK, null,
null);
+ } else {
+ LOG.warn("Unable to read the ledger: {} information",
ledgerId);
+
iterCallback.processResult(BKException.getExceptionCode(exception), null, null);
+ }
+ });
+ }
+ };
+ // Reading the result after processing all the ledgers
+ final List<Integer> resultCode = new ArrayList<Integer>(1);
+ ledgerManager.asyncProcessLedgers(ledgerProcessor, new
AsyncCallback.VoidCallback() {
+
+ @Override
+ public void processResult(int rc, String s, Object obj) {
+ resultCode.add(rc);
+ metadataCheckLatch.countDown();
+ }
+ }, null, BKException.Code.OK, BKException.Code.ReadException);
+ try {
+ metadataCheckLatch.await();
Review comment:
We could put an upperbound to the ovetall processing.
The downside is that ledgers with greater id could starve
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
