You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@bookkeeper.apache.org by GitBox <gi...@apache.org> on 2018/02/19 07:54:30 UTC
[GitHub] sijie closed pull request #1140: Issue 1139: Add debug to replication fencing
sijie closed pull request #1140: Issue 1139: Add debug to replication fencing
URL: https://github.com/apache/bookkeeper/pull/1140
This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:
As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):
diff --git a/bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/ReplicationWorker.java b/bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/ReplicationWorker.java
index 65c0c2ce2..f6ed7d1ec 100644
--- a/bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/ReplicationWorker.java
+++ b/bookkeeper-server/src/main/java/org/apache/bookkeeper/replication/ReplicationWorker.java
@@ -339,6 +339,10 @@ private boolean isLastSegmentOpenAndMissingBookies(LedgerHandle lh) throws BKExc
Collection<BookieSocketAddress> available = admin.getAvailableBookies();
for (BookieSocketAddress b : finalEnsemble) {
if (!available.contains(b)) {
+ if (LOG.isDebugEnabled()) {
+ LOG.debug("Bookie {} is missing from the list of Available Bookies. ledger {}:ensemble {}.",
+ b, lh.getId(), finalEnsemble);
+ }
return true;
}
}
@@ -366,32 +370,45 @@ private void deferLedgerLockRelease(final long ledgerId) {
TimerTask timerTask = new TimerTask() {
@Override
public void run() {
+ boolean isRecoveryOpen = false;
LedgerHandle lh = null;
try {
lh = admin.openLedgerNoRecovery(ledgerId);
if (isLastSegmentOpenAndMissingBookies(lh)) {
+ // Need recovery open, close the old ledger handle.
+ lh.close();
+ // Recovery open could result in client write failure.
+ LOG.warn("Missing bookie(s) from last segment. Opening Ledger{} for Recovery.", ledgerId);
lh = admin.openLedger(ledgerId);
+ isRecoveryOpen = true;
}
-
- Set<LedgerFragment> fragments =
- getUnderreplicatedFragments(lh, conf.getAuditorLedgerVerificationPercentage());
- for (LedgerFragment fragment : fragments) {
- if (!fragment.isClosed()) {
- lh = admin.openLedger(ledgerId);
- break;
+ if (!isRecoveryOpen){
+ Set<LedgerFragment> fragments =
+ getUnderreplicatedFragments(lh, conf.getAuditorLedgerVerificationPercentage());
+ for (LedgerFragment fragment : fragments) {
+ if (!fragment.isClosed()) {
+ // Need recovery open, close the old ledger handle.
+ lh.close();
+ // Recovery open could result in client write failure.
+ LOG.warn("Open Fragment{}. Opening Ledger{} for Recovery.",
+ fragment.getEnsemble(), ledgerId);
+ lh = admin.openLedger(ledgerId);
+ isRecoveryOpen = true;
+ break;
+ }
}
}
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
- LOG.info("InterruptedException "
- + "while replicating fragments", e);
+ LOG.info("InterruptedException while fencing the ledger {}"
+ + " for rereplication of postponed ledgers", ledgerId, e);
} catch (BKNoSuchLedgerExistsException bknsle) {
if (LOG.isDebugEnabled()) {
- LOG.debug("Ledger was deleted, safe to continue", bknsle);
+ LOG.debug("Ledger {} was deleted, safe to continue", ledgerId, bknsle);
}
} catch (BKException e) {
- LOG.error("BKException while fencing the ledger"
- + " for rereplication of postponed ledgers", e);
+ LOG.error("BKException while fencing the ledger {}"
+ + " for rereplication of postponed ledgers", ledgerId, e);
} finally {
try {
if (lh != null) {
@@ -399,20 +416,19 @@ public void run() {
}
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
- LOG.info("InterruptedException while closing "
- + "ledger", e);
+ LOG.info("InterruptedException while closing ledger {}", ledgerId, e);
} catch (BKException e) {
// Lets go ahead and release the lock. Catch actual
// exception in normal replication flow and take
// action.
- LOG.warn("BKException while closing ledger ", e);
+ LOG.warn("BKException while closing ledger {} ", ledgerId, e);
} finally {
try {
underreplicationManager
.releaseUnderreplicatedLedger(ledgerId);
} catch (UnavailableException e) {
- LOG.error("UnavailableException "
- + "while replicating fragments", e);
+ LOG.error("UnavailableException while replicating fragments of ledger {}",
+ ledgerId, e);
shutdown();
}
}
----------------------------------------------------------------
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:
users@infra.apache.org
With regards,
Apache Git Services