[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[ https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16181312#comment-16181312 ] ASF subversion and git services commented on ARTEMIS-1417: -- Commit f51023506e20101cea87676df44337a019e24cc6 in activemq-artemis's branch refs/heads/1.x from [~jbertram] [ https://git-wip-us.apache.org/repos/asf?p=activemq-artemis.git;h=f510235 ] ARTEMIS-1417 Failback not working on NFSv4 With NFSv4 it is now necessary to lock/unlock the byte of the server lock file where the state information is written so that the information is then flushed to the other clients looking at the file. (cherry picked from commit 2ec173bc708daca163c9356cf12440432abc61c4) > Failback not working on NFSv4 > - > > Key: ARTEMIS-1417 > URL: https://issues.apache.org/jira/browse/ARTEMIS-1417 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: Broker >Affects Versions: 2.3.0 >Reporter: Justin Bertram >Assignee: Justin Bertram > Fix For: 2.4.0 > > > If Artemis is using NFSv4 as shared store for its messaging journal to > provide HA then a write to server.lock made by one server may not be > propagated to the other server. This prevents failback from happening. This > is change in behavior in data synchronization between two NFSv4 clients was > observed when recently upgrading Linux platforms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[ https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16164471#comment-16164471 ] ASF GitHub Bot commented on ARTEMIS-1417: - Github user asfgit closed the pull request at: https://github.com/apache/activemq-artemis/pull/1530 > Failback not working on NFSv4 > - > > Key: ARTEMIS-1417 > URL: https://issues.apache.org/jira/browse/ARTEMIS-1417 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: Broker >Affects Versions: 2.3.0 >Reporter: Justin Bertram >Assignee: Justin Bertram > Fix For: 2.4.0 > > > If Artemis is using NFSv4 as shared store for its messaging journal to > provide HA then a write to server.lock made by one server may not be > propagated to the other server. This prevents failback from happening. This > is change in behavior in data synchronization between two NFSv4 clients was > observed when recently upgrading Linux platforms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[ https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16164470#comment-16164470 ] ASF subversion and git services commented on ARTEMIS-1417: -- Commit 2ec173bc708daca163c9356cf12440432abc61c4 in activemq-artemis's branch refs/heads/master from [~jbertram] [ https://git-wip-us.apache.org/repos/asf?p=activemq-artemis.git;h=2ec173b ] ARTEMIS-1417 Failback not working on NFSv4 With NFSv4 it is now necessary to lock/unlock the byte of the server lock file where the state information is written so that the information is then flushed to the other clients looking at the file. > Failback not working on NFSv4 > - > > Key: ARTEMIS-1417 > URL: https://issues.apache.org/jira/browse/ARTEMIS-1417 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: Broker >Affects Versions: 2.3.0 >Reporter: Justin Bertram >Assignee: Justin Bertram > Fix For: 2.4.0 > > > If Artemis is using NFSv4 as shared store for its messaging journal to > provide HA then a write to server.lock made by one server may not be > propagated to the other server. This prevents failback from happening. This > is change in behavior in data synchronization between two NFSv4 clients was > observed when recently upgrading Linux platforms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[
https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16164468#comment-16164468
]
ASF GitHub Bot commented on ARTEMIS-1417:
-
Github user mtaylor commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/1530#discussion_r138584661
--- Diff:
artemis-server/src/main/java/org/apache/activemq/artemis/core/server/impl/FileLockNodeManager.java
---
@@ -267,25 +293,27 @@ public final SimpleString readNodeId() throws
ActiveMQIllegalStateException, IOE
return getNodeId();
}
- protected FileLock tryLock(final int lockPos) throws Exception {
+ protected FileLock tryLock(final long lockPos) throws IOException {
--- End diff --
+1 on this. Merging.
> Failback not working on NFSv4
> -
>
> Key: ARTEMIS-1417
> URL: https://issues.apache.org/jira/browse/ARTEMIS-1417
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
>Affects Versions: 2.3.0
>Reporter: Justin Bertram
>Assignee: Justin Bertram
> Fix For: 2.4.0
>
>
> If Artemis is using NFSv4 as shared store for its messaging journal to
> provide HA then a write to server.lock made by one server may not be
> propagated to the other server. This prevents failback from happening. This
> is change in behavior in data synchronization between two NFSv4 clients was
> observed when recently upgrading Linux platforms.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[
https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16163263#comment-16163263
]
ASF GitHub Bot commented on ARTEMIS-1417:
-
Github user michaelandrepearce commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/1530#discussion_r138404539
--- Diff:
artemis-server/src/main/java/org/apache/activemq/artemis/core/server/impl/FileLockNodeManager.java
---
@@ -267,25 +293,27 @@ public final SimpleString readNodeId() throws
ActiveMQIllegalStateException, IOE
return getNodeId();
}
- protected FileLock tryLock(final int lockPos) throws Exception {
+ protected FileLock tryLock(final long lockPos) throws IOException {
--- End diff --
thanks for explanation, and +1 for consistency cheers.
> Failback not working on NFSv4
> -
>
> Key: ARTEMIS-1417
> URL: https://issues.apache.org/jira/browse/ARTEMIS-1417
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
>Affects Versions: 2.3.0
>Reporter: Justin Bertram
>Assignee: Justin Bertram
> Fix For: 2.4.0
>
>
> If Artemis is using NFSv4 as shared store for its messaging journal to
> provide HA then a write to server.lock made by one server may not be
> propagated to the other server. This prevents failback from happening. This
> is change in behavior in data synchronization between two NFSv4 clients was
> observed when recently upgrading Linux platforms.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[
https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16163261#comment-16163261
]
ASF GitHub Bot commented on ARTEMIS-1417:
-
Github user jbertram commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/1530#discussion_r138404056
--- Diff:
artemis-server/src/main/java/org/apache/activemq/artemis/core/server/impl/FileLockNodeManager.java
---
@@ -267,25 +293,27 @@ public final SimpleString readNodeId() throws
ActiveMQIllegalStateException, IOE
return getNodeId();
}
- protected FileLock tryLock(final int lockPos) throws Exception {
+ protected FileLock tryLock(final long lockPos) throws IOException {
--- End diff --
java.nio.channels.FileChannel#tryLock takes longs so that's why I changed
it. I updated the rest of the methods/variables to be consistent.
> Failback not working on NFSv4
> -
>
> Key: ARTEMIS-1417
> URL: https://issues.apache.org/jira/browse/ARTEMIS-1417
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
>Affects Versions: 2.3.0
>Reporter: Justin Bertram
>Assignee: Justin Bertram
> Fix For: 2.4.0
>
>
> If Artemis is using NFSv4 as shared store for its messaging journal to
> provide HA then a write to server.lock made by one server may not be
> propagated to the other server. This prevents failback from happening. This
> is change in behavior in data synchronization between two NFSv4 clients was
> observed when recently upgrading Linux platforms.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[
https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16163257#comment-16163257
]
ASF GitHub Bot commented on ARTEMIS-1417:
-
Github user michaelandrepearce commented on a diff in the pull request:
https://github.com/apache/activemq-artemis/pull/1530#discussion_r138402896
--- Diff:
artemis-server/src/main/java/org/apache/activemq/artemis/core/server/impl/FileLockNodeManager.java
---
@@ -267,25 +293,27 @@ public final SimpleString readNodeId() throws
ActiveMQIllegalStateException, IOE
return getNodeId();
}
- protected FileLock tryLock(final int lockPos) throws Exception {
+ protected FileLock tryLock(final long lockPos) throws IOException {
--- End diff --
lock is changed from int to long here, was there a need?
if so other methods such as lock should be updated, if not this should be
reverted
> Failback not working on NFSv4
> -
>
> Key: ARTEMIS-1417
> URL: https://issues.apache.org/jira/browse/ARTEMIS-1417
> Project: ActiveMQ Artemis
> Issue Type: Bug
> Components: Broker
>Affects Versions: 2.3.0
>Reporter: Justin Bertram
>Assignee: Justin Bertram
> Fix For: 2.4.0
>
>
> If Artemis is using NFSv4 as shared store for its messaging journal to
> provide HA then a write to server.lock made by one server may not be
> propagated to the other server. This prevents failback from happening. This
> is change in behavior in data synchronization between two NFSv4 clients was
> observed when recently upgrading Linux platforms.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[ https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16163198#comment-16163198 ] Gary Tully commented on ARTEMIS-1417: - There may be some inspiration in 5.x - we have taken to writing to the lockfile to ensure it has not been deleted and recreated. see some history at: https://github.com/apache/activemq/commits/57795bafcea290c6879bb288822435c480a9212d/activemq-broker/src/main/java/org/apache/activemq/util/LockFile.java > Failback not working on NFSv4 > - > > Key: ARTEMIS-1417 > URL: https://issues.apache.org/jira/browse/ARTEMIS-1417 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: Broker >Affects Versions: 2.3.0 >Reporter: Justin Bertram >Assignee: Justin Bertram > Fix For: 2.4.0 > > > If Artemis is using NFSv4 as shared store for its messaging journal to > provide HA then a write to server.lock made by one server may not be > propagated to the other server. This prevents failback from happening. This > is change in behavior in data synchronization between two NFSv4 clients was > observed when recently upgrading Linux platforms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
[jira] [Commented] (ARTEMIS-1417) Failback not working on NFSv4
[ https://issues.apache.org/jira/browse/ARTEMIS-1417?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16163196#comment-16163196 ] ASF GitHub Bot commented on ARTEMIS-1417: - GitHub user jbertram opened a pull request: https://github.com/apache/activemq-artemis/pull/1530 ARTEMIS-1417 Failback not working on NFSv4 With NFSv4 it is now necessary to lock/unlock the byte of the server lock file where the state information is written so that the information is then flushed to the other clients looking at the file. You can merge this pull request into a Git repository by running: $ git pull https://github.com/jbertram/activemq-artemis ARTEMIS-1417 Alternatively you can review and apply these changes as the patch at: https://github.com/apache/activemq-artemis/pull/1530.patch To close this pull request, make a commit to your master/trunk branch with (at least) the following in the commit message: This closes #1530 commit d52828ed6a59f16f94e5ef1445b4b7b28802b81b Author: Justin Bertram Date: 2017-09-12T16:17:08Z ARTEMIS-1417 Failback not working on NFSv4 With NFSv4 it is now necessary to lock/unlock the byte of the server lock file where the state information is written so that the information is then flushed to the other clients looking at the file. > Failback not working on NFSv4 > - > > Key: ARTEMIS-1417 > URL: https://issues.apache.org/jira/browse/ARTEMIS-1417 > Project: ActiveMQ Artemis > Issue Type: Bug > Components: Broker >Affects Versions: 2.3.0 >Reporter: Justin Bertram >Assignee: Justin Bertram > Fix For: 2.4.0 > > > If Artemis is using NFSv4 as shared store for its messaging journal to > provide HA then a write to server.lock made by one server may not be > propagated to the other server. This prevents failback from happening. This > is change in behavior in data synchronization between two NFSv4 clients was > observed when recently upgrading Linux platforms. -- This message was sent by Atlassian JIRA (v6.4.14#64029)
