dlmarion commented on code in PR #5070:
URL: https://github.com/apache/accumulo/pull/5070#discussion_r1844212773
##########
server/manager/src/main/java/org/apache/accumulo/manager/Manager.java:
##########
@@ -1057,11 +1059,29 @@ private long balanceTablets() {
int attemptNum = 0;
do {
log.debug("Balancing for tables at level {}, times-in-loop: {}", dl,
++attemptNum);
- params = BalanceParamsImpl.fromThrift(tserverStatusForBalancerLevel,
- tserverStatusForLevel, partitionedMigrations.get(dl));
+
+ SortedMap<TabletServerId,TServerStatus> statusForBalancerLevel =
+ tserverStatusForBalancerLevel;
+ if (attemptNum % checkInterval == 0) {
Review Comment:
Even though we are passing the balancer a set of tablets for the root or
metadata table, that doesn't prevent it from returning migrations for other
tables. I think the HostRegexTableLoadBalancer will do this
[here](https://github.com/apache/accumulo/blob/2.1/core/src/main/java/org/apache/accumulo/core/spi/balancer/HostRegexTableLoadBalancer.java#L400).
In this case, we could be operating on the root or metadata level with
`migrationsOutForLevel` being greater than zero. If we exit and call this
method again, then I think we end up at the same place.
--
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]
