EdColeman commented on code in PR #3230:
URL: https://github.com/apache/accumulo/pull/3230#discussion_r1140585817
##########
server/manager/src/main/java/org/apache/accumulo/manager/upgrade/PreUpgradeValidation.java:
##########
@@ -117,4 +138,111 @@ private void fail(Exception e) {
System.exit(1);
}
+ private void validateTableLocks(final ServerContext context) {
+
+ final ZooReaderWriter zrw = context.getZooReaderWriter();
+ final ZooKeeper zk = zrw.getZooKeeper();
+ final String rootPath = context.getZooKeeperRoot();
+ final String tserverLockRoot = rootPath + Constants.ZTSERVERS;
+
+ log.debug("Looking for locks that may be from previous version in path:
{}", tserverLockRoot);
+
+ AtomicInteger errorCount = new AtomicInteger(0);
+
+ List<Pair<HostAndPort,ServiceLock.ServiceLockPath>> hostsWithLocks =
+ gatherLocks(zk, tserverLockRoot);
+
+ // try a thrift call to the hosts - hosts running previous versions will
fail the call
+ ThreadPoolExecutor lockCheckPool =
ThreadPools.getServerThreadPools().createThreadPool(8, 64,
+ 10, MINUTES, "update-lock-check", false);
+
+ hostsWithLocks.forEach(h -> lockCheckPool.execute(() -> {
+ HostAndPort host = h.getFirst();
+ ServiceLock.ServiceLockPath lockPath = h.getSecond();
+ try (TTransport transport = ThriftUtil.createTransport(host, context)) {
+ log.trace("found valid lock at: {}", lockPath);
+ } catch (TException ex) {
+ log.debug("Could not establish a connection for to service holding
lock. Deleting node: {}",
+ lockPath, ex);
+ try {
+ zk.delete(lockPath.toString(), -1);
+ errorCount.incrementAndGet();
+ } catch (KeeperException.NoNodeException e) {
+ // ignore - node already gone.
+ } catch (InterruptedException | KeeperException e) {
+ // task will be terminated - ignore interrupt.
+ errorCount.incrementAndGet();
+ }
+ }
+ }));
+ lockCheckPool.shutdown();
+ try {
+ // wait to all to finish
+ if (!lockCheckPool.awaitTermination(10, MINUTES)) {
+ log.warn(
+ "Timed out waiting for lock check to finish - continuing, but
tservers running prior versions may be present");
+ }
+ } catch (InterruptedException ex) {
+ Thread.currentThread().interrupt();
Review Comment:
Reworked in a1ff3fa6ee6c6342
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
