[GitHub] [accumulo] ctubbsii commented on issue #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList

2019-09-12 Thread GitBox
ctubbsii commented on issue #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList URL: https://github.com/apache/accumulo/pull/1333#issuecomment-530684737 @belugabehr I'm not going to be able to continue following the issue this week, because I'm at ApacheCon right now and

[GitHub] [accumulo] belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530824196 @ctubbsii Yes. Tablet appears to be a 'hot' lock. I am actually looking at that class but it's a bit of a beast. Anyway we can

[GitHub] [accumulo] keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530861250 > Update to use Java Concurrency classes (as an example for the rest of the code base) I think this could be a nice

[GitHub] [accumulo] belugabehr commented on a change in pull request #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList

2019-09-12 Thread GitBox
belugabehr commented on a change in pull request #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList URL: https://github.com/apache/accumulo/pull/1333#discussion_r323728667 ## File path:

[GitHub] [accumulo] keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530859174 The code is very complicated for Tablet, but the conceptual model is very simple. There is a single lock that protects most

[GitHub] [accumulo] keith-turner edited a comment on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
keith-turner edited a comment on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530861250 > Update to use Java Concurrency classes (as an example for the rest of the code base) I think this could be a

[GitHub] [accumulo] belugabehr commented on a change in pull request #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList

2019-09-12 Thread GitBox
belugabehr commented on a change in pull request #1333: Removed LinkedList from CachedBlockQueue in favor of ArrayList URL: https://github.com/apache/accumulo/pull/1333#discussion_r323728667 ## File path:

[GitHub] [accumulo] keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#discussion_r323832463 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/RowLocks.java ##

[GitHub] [accumulo] keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#discussion_r323823547 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/RowLocks.java ##

[GitHub] [accumulo-proxy] keith-turner opened a new issue #7: Look into using Accumulo maven plugin for Proxy test

2019-09-12 Thread GitBox
keith-turner opened a new issue #7: Look into using Accumulo maven plugin for Proxy test URL: https://github.com/apache/accumulo-proxy/issues/7 Currently the proxy uses internal Accumulo code for ITs. It could use the [Accumulo maven

[GitHub] [accumulo] belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530933558 @keith-turner Thanks for the input. So, the downside with your proposed solution is that the locking becomes very

[GitHub] [accumulo] hkeebler commented on a change in pull request #1330: Fix #1090 Consolidate/Simplfy ServerOpts

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1330: Fix #1090 Consolidate/Simplfy ServerOpts URL: https://github.com/apache/accumulo/pull/1330#discussion_r323881903 ## File path: server/gc/src/main/java/org/apache/accumulo/gc/SimpleGarbageCollector.java ## @@ -258,12

[GitHub] [accumulo] belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530936898 Also, typically, I use `isHeldByCurrentThread()` in `private` scoped methods to ensure the lock is not lost with a future

[GitHub] [accumulo-proxy] keith-turner commented on issue #6: Ensure proxy only uses public API before release

2019-09-12 Thread GitBox
keith-turner commented on issue #6: Ensure proxy only uses public API before release URL: https://github.com/apache/accumulo-proxy/issues/6#issuecomment-530953243 Issue number #7 mentions ITs use non public Accumulo APIs. Personally I think its ok if test code is still using non-public

[GitHub] [accumulo-proxy] keith-turner opened a new issue #6: Ensure proxy only uses public API before release

2019-09-12 Thread GitBox
keith-turner opened a new issue #6: Ensure proxy only uses public API before release URL: https://github.com/apache/accumulo-proxy/issues/6 When the proxy was released with Accumulo it was ok if used non-public API Accumulo code because it was always released with Accumulo. Now that it

[GitHub] [accumulo] EdColeman commented on issue #1352: Fix #1345 - Allow gc initiated compaction as option.

2019-09-12 Thread GitBox
EdColeman commented on issue #1352: Fix #1345 - Allow gc initiated compaction as option. URL: https://github.com/apache/accumulo/pull/1352#issuecomment-530941370 I'm going to merge this unless there are objections. When I merge forward I'll change the default from compact to flush.

[GitHub] [accumulo] belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530944412 You can check out my work on the matter here: https://github.com/belugabehr/accumulo/tree/TabletLock

[GitHub] [accumulo-proxy] keith-turner opened a new issue #8: Ensure proxy works well in docker

2019-09-12 Thread GitBox
keith-turner opened a new issue #8: Ensure proxy works well in docker URL: https://github.com/apache/accumulo-proxy/issues/8 Not sure all that this entails. When making Fluo and Accumulo work well in docker it was important that the script could accept options on the command line. Would

[GitHub] [accumulo] keith-turner commented on issue #995: Look into using array of locks for conditional mutation row locks

2019-09-12 Thread GitBox
keith-turner commented on issue #995: Look into using array of locks for conditional mutation row locks URL: https://github.com/apache/accumulo/issues/995#issuecomment-530906586 Superseded by #1357 This is an automated

[GitHub] [accumulo] keith-turner closed issue #995: Look into using array of locks for conditional mutation row locks

2019-09-12 Thread GitBox
keith-turner closed issue #995: Look into using array of locks for conditional mutation row locks URL: https://github.com/apache/accumulo/issues/995 This is an automated message from the Apache Git Service. To respond to

[GitHub] [accumulo] hkeebler commented on issue #1330: Fix #1090 Consolidate/Simplfy ServerOpts

2019-09-12 Thread GitBox
hkeebler commented on issue #1330: Fix #1090 Consolidate/Simplfy ServerOpts URL: https://github.com/apache/accumulo/pull/1330#issuecomment-530943813 Last commit includes all suggested changes. This is an automated message

[GitHub] [accumulo] keith-turner edited a comment on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
keith-turner edited a comment on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530944944 > So, the downside with your proposed solution is that the locking becomes very granular. That is to say, if a method

[GitHub] [accumulo] keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
keith-turner commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-530944944 > So, the downside with your proposed solution is that the locking becomes very granular. That is to say, if a method needs to

Accumulo-1.9 - Build # 411 - Fixed

2019-09-12 Thread Apache Jenkins Server
The Apache Jenkins build system has built Accumulo-1.9 (build #411) Status: Fixed Check console output at https://builds.apache.org/job/Accumulo-1.9/411/ to view the results.

[GitHub] [accumulo] belugabehr commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
belugabehr commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#issuecomment-531001758 @keith-turner Teamwork makes the dream work. Love it. Have at it.

[GitHub] [accumulo] keith-turner commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#issuecomment-531009014 I ran ConditionalWriterIT which exercises RowLocks and it passed.

[GitHub] [accumulo] belugabehr commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
belugabehr commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#discussion_r323887724 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/RowLocks.java ## @@

[GitHub] [accumulo] EdColeman opened a new pull request #1360: Improve concurrent access to gc statistics.

2019-09-12 Thread GitBox
EdColeman opened a new pull request #1360: Improve concurrent access to gc statistics. URL: https://github.com/apache/accumulo/pull/1360 This pull request is intended to allow safe, concurrent access to the gc metrics so they can be accessed by the metrics system. The ultimate goal is to

[GitHub] [accumulo-proxy] keith-turner commented on issue #8: Ensure proxy works well in docker

2019-09-12 Thread GitBox
keith-turner commented on issue #8: Ensure proxy works well in docker URL: https://github.com/apache/accumulo-proxy/issues/8#issuecomment-530958384 To be more precise I am thinking the tarball proposed in #5 should work well when built into a docker container.

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323910821 ## File path: core/src/test/java/org/apache/accumulo/core/metadata/schema/SortSkewTest.java ## @@ -0,0

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323910545 ## File path: server/base/src/main/java/org/apache/accumulo/server/metadata/ServerAmpleImpl.java ## @@

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323917658 ## File path: test/src/main/java/org/apache/accumulo/test/functional/GarbageCollectorIT.java ## @@

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323917658 ## File path: test/src/main/java/org/apache/accumulo/test/functional/GarbageCollectorIT.java ## @@

[GitHub] [accumulo] EdColeman closed issue #1345: GC forces metadata compaction every cycle

2019-09-12 Thread GitBox
EdColeman closed issue #1345: GC forces metadata compaction every cycle URL: https://github.com/apache/accumulo/issues/1345 This is an automated message from the Apache Git Service. To respond to the message, please log on

[GitHub] [accumulo] keith-turner merged pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner merged pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357 This is an automated message from the Apache Git Service. To respond to the

[GitHub] [accumulo] keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#discussion_r323942654 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/RowLocks.java ##

[GitHub] [accumulo-testing] milleruntime opened a new issue #111: AssertionError thrown where never should happen

2019-09-12 Thread GitBox
milleruntime opened a new issue #111: AssertionError thrown where never should happen URL: https://github.com/apache/accumulo-testing/issues/111 I saw this while running a rwalk on a 5 node cluster. ``` 2019-09-12 21:26:51,842 [testing.randomwalk.Module] ERROR: Caught error

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323900617 ## File path: core/src/main/java/org/apache/accumulo/core/metadata/schema/SortSkew.java ## @@ -0,0

[GitHub] [accumulo] EdColeman merged pull request #1352: Fix #1345 - Allow gc initiated compaction as option.

2019-09-12 Thread GitBox
EdColeman merged pull request #1352: Fix #1345 - Allow gc initiated compaction as option. URL: https://github.com/apache/accumulo/pull/1352 This is an automated message from the Apache Git Service. To respond to the

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323911245 ## File path: test/src/main/java/org/apache/accumulo/test/functional/GarbageCollectorIT.java ## @@

[GitHub] [accumulo] hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points

2019-09-12 Thread GitBox
hkeebler commented on a change in pull request #1344: Fix #1043 Support stable ~del split points URL: https://github.com/apache/accumulo/pull/1344#discussion_r323911541 ## File path: server/base/src/main/java/org/apache/accumulo/server/util/ListVolumesUsed.java ## @@

[GitHub] [accumulo] keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on a change in pull request #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#discussion_r323941807 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/RowLocks.java ##

[GitHub] [accumulo-proxy] keith-turner edited a comment on issue #8: Ensure proxy works well in docker

2019-09-12 Thread GitBox
keith-turner edited a comment on issue #8: Ensure proxy works well in docker URL: https://github.com/apache/accumulo-proxy/issues/8#issuecomment-530958384 To be more precise I am thinking the tarball proposed in #5 should work well when used by a docker build file to create a docker image.

[GitHub] [accumulo-proxy] keith-turner edited a comment on issue #8: Ensure proxy works well in docker

2019-09-12 Thread GitBox
keith-turner edited a comment on issue #8: Ensure proxy works well in docker URL: https://github.com/apache/accumulo-proxy/issues/8#issuecomment-530958384 To be more precise I am thinking the tarball proposed in #5 should work well when used by a simple docker build file to create a docker

[GitHub] [accumulo] belugabehr opened a new pull request #1359: Use Java Conccurent to manage Timeouts

2019-09-12 Thread GitBox
belugabehr opened a new pull request #1359: Use Java Conccurent to manage Timeouts URL: https://github.com/apache/accumulo/pull/1359 Current code does not handle a "spurious wakeup":

[GitHub] [accumulo] keith-turner commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks

2019-09-12 Thread GitBox
keith-turner commented on issue #1357: Use ConcurrentHashMap instead of synchronized blocks URL: https://github.com/apache/accumulo/pull/1357#issuecomment-531008392 I pushed a small change a3107c08a82e2b6a209ca70324c35c5573e4c385. When I was looking at this PR I noticed one of the

Accumulo-Master - Build # 2924 - Aborted

2019-09-12 Thread Apache Jenkins Server
The Apache Jenkins build system has built Accumulo-Master (build #2924) Status: Aborted Check console output at https://builds.apache.org/job/Accumulo-Master/2924/ to view the results.

[GitHub] [accumulo] belugabehr commented on a change in pull request #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on a change in pull request #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#discussion_r323984963 ## File path: server/tserver/src/main/java/org/apache/accumulo/tserver/tablet/CommitSession.java

[GitHub] [accumulo] belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter

2019-09-12 Thread GitBox
belugabehr commented on issue #1358: Fix race conditions with commitsInProgress counter URL: https://github.com/apache/accumulo/pull/1358#issuecomment-531040596 For this particular PR it's still not clear to me why `CommitSession.java` needs to lock (block) the entire tablet to keep this