Re: [Artemis] upgrade from 2.1.0 getting errors around config-delete-queues

2018-07-21 Thread Dan Langford
in case its helpful ill add that when i look at the move from 2.1.0 to 2.2.0 https://github.com/apache/activemq-artemis/compare/2.1.0...2.2.0 i can go to "Files changed", find AddressSettings.java and on line 724 (on the 2.2.0 side) i can see the broker trying to decode a policyStr and it can allow

Re: [Artemis] upgrade from 2.1.0 getting errors around config-delete-queues

2018-07-21 Thread Dan Langford
I am to blame for introducing the confusion across threads. Sorry I know that when I am on 2.1.0 and I have Queues and address settings defined via the API at runtime (so they are not defined in the broker.xml) I absolutely cannot upgrade from 2.1.0 to 2.2.0, 2.3.0, 2.4.0, or 2.5.0. So the issue m

Re: [Artemis] upgrade from 2.1.0 getting errors around config-delete-queues

2018-07-21 Thread Clebert Suconic
Ouch. I crossed posts. I thought this was the 30 min load thing. Still u sure this is not fixed beyond 2.1.0? On Sat, Jul 21, 2018 at 2:26 PM Clebert Suconic wrote: > @Justin why open a new JIRA? I’m pretty sure this is an issue I fixed > beyond 2.2.0. > > So unless this proves to be a new i

Re: [Artemis] upgrade from 2.1.0 getting errors around config-delete-queues

2018-07-21 Thread Clebert Suconic
@Justin why open a new JIRA? I’m pretty sure this is an issue I fixed beyond 2.2.0. So unless this proves to be a new issue. Just have the user trying the new version ? On Fri, Jul 20, 2018 at 6:00 PM Justin Bertram wrote: > I'm looking at a possible solution to this issue. Please open a JIR

Re: Using AWS EFS as the shared file system for master/slave broker pair

2018-07-21 Thread Tim Bain
Third reply, hopefully this is the last thing I missed. If pluggable lockers are required in order to use EFS, they would be required irrespective of which version of ActiveMQ you use. However, the EFS FAQ says that EFS supports NFS 4.1 locking semantics, which sounds to me like pluggable lockers

Re: Using AWS EFS as the shared file system for master/slave broker pair

2018-07-21 Thread Tim Bain
I forgot to answer your other question. That just means that you have to configure the path for the KahaDB data directory (in activemq.xml) to be a path under your EFS mount. Tim On Sat, Jul 21, 2018, 7:17 AM Tim Bain wrote: > The bug I linked to, which addresses the fact that EFS is so large

Re: Using AWS EFS as the shared file system for master/slave broker pair

2018-07-21 Thread Tim Bain
The bug I linked to, which addresses the fact that EFS is so large that the ActiveMQ code believes it has a negative size, was fixed in 5.15.0, so I would expect 5.14.3 not to work. But YMMV if EFS is reporting a different amount of free space today than it did when that bug was submitted. Tim On