Re: [Artemis] Master fails to start up after failback

On Thu, Oct 22, 2015 at 4:22 AM, Mihkel Nõges wrote: > Hi Martyn, Clebert, > > Thanks for fixing it and for good suggestions! I think the best way still > is to not use failback in production at all and instead change the master > configuration to be slave after the master crashes or is stopped an

Re: [Artemis] Master fails to start up after failback

Hi Martyn, Clebert, Thanks for fixing it and for good suggestions! I think the best way still is to not use failback in production at all and instead change the master configuration to be slave after the master crashes or is stopped and needs to be restarted. I guess this is the more safer way of

Re: [Artemis] Master fails to start up after failback

another possible workaround is to start the server from where your paths are relative from. Or you could try the snapshot build: https://repository.apache.org/content/repositories/snapshots/org/apache/activemq/apache-artemis/1.1.1-SNAPSHOT/apache-artemis-1.1.1-20151021.162952-18-bin.zip On Wed,

Re: [Artemis] Master fails to start up after failback

This should now be fixed upstream as part of: https://issues.apache.org/jira/browse/ARTEMIS-273 On 20/10/15 19:29, Mihkel Nõges wrote: Thanks Martyn! I will try this tomorrow. BR! Mihkel On 20 October 2015 at 18:59, Martyn Taylor wrote: Hi Mihkel, I tried reproducing this locally and ran

Re: [Artemis] Master fails to start up after failback

Thanks Martyn! I will try this tomorrow. BR! Mihkel On 20 October 2015 at 18:59, Martyn Taylor wrote: > Hi Mihkel, > > I tried reproducing this locally and ran into an issue straight away when > running the example. After some investigation it appears that there is an > issue in the ArtemisSe

Re: [Artemis] Master fails to start up after failback

eplication as the journal contains the unique ID of each node. > > > Justin > > - Original Message - > From: "Mihkel Nõges" > To: users@activemq.apache.org > Sent: Tuesday, October 20, 2015 9:46:24 AM > Subject: Re: [Artemis] Master fails to start up after

Re: [Artemis] Master fails to start up after failback

Hi Mihkel, I tried reproducing this locally and ran into an issue straight away when running the example. After some investigation it appears that there is an issue in the ArtemisServerImpl which is preventing the journal files from being replicated properly when using relative paths in the

Re: [Artemis] Master fails to start up after failback

replication as the journal contains the unique ID of each node. Justin - Original Message - From: "Mihkel Nõges" To: users@activemq.apache.org Sent: Tuesday, October 20, 2015 9:46:24 AM Subject: Re: [Artemis] Master fails to start up after failback Yes, sorry, had typo in e

Re: [Artemis] Master fails to start up after failback

gt; > > Justin > > - Original Message - > From: "Mihkel Nõges" > To: users@activemq.apache.org > Sent: Tuesday, October 20, 2015 9:24:21 AM > Subject: Re: [Artemis] Master fails to start up after failback > > Also I had a question earlier about having more tha

Re: [Artemis] Master fails to start up after failback

You can't have 2 masters using the same shared-store. However, you can have 2 masters each with their own store. Justin - Original Message - From: "Mihkel Nõges" To: users@activemq.apache.org Sent: Tuesday, October 20, 2015 9:24:21 AM Subject: Re: [Artemis] Master fa

Re: [Artemis] Master fails to start up after failback

Sorry, the last mail went out too fast. Instead of shared-store I had replication On 20 October 2015 at 17:24, Mihkel Nõges wrote: > Also I had a question earlier about having more than one Artemis master in > single cluster. When I tried this it resulted in only one master becoming a > master,

Re: [Artemis] Master fails to start up after failback

Also I had a question earlier about having more than one Artemis master in single cluster. When I tried this it resulted in only one master becoming a master, the other one became a slave for the first one started even though I set different group-name values for them in broker.xml. Is this expecte

Re: [Artemis] Master fails to start up after failback

Hi Tim, Clebert! Yes we considered also the alternatives ( http://activemq.apache.org/masterslave.html): *Shared Storage:* We do not have high performance shared storage solution. We have some solution for our current file storage needs, but it's I/O is said to be very slow and would need to be e

Re: [Artemis] Master fails to start up after failback

We Artemis, there is the shared storage (same as with KahaDB) and replication. We were talking specifically about replication here... you would have a single node using the file system. It's always my preferred choice for a production system actually. (shared storage). No matter AMQ5 or Artemis.

Re: [Artemis] Master fails to start up after failback

ActiveMQ doesn't have any process for letting the original master take back over, though you could achieve it by issuing a restart command to the original slave broker when you bring the original master back up (and you could script that easily if you wanted to). ActiveMQ does sync the data betwee

Re: [Artemis] Master fails to start up after failback

As far as I know ActiveMQ5 doesn't do failback on the master-slave journal... and it doesn't have any protocol to sync the data between master and slave. There is a small regression on the failback that we are dealing now... if you set the master as a backup it would work fine... I think your t

Re: [Artemis] Master fails to start up after failback

If you were to move to ActiveMQ, whyis JDBC your message store of choice? The other two options have better performance, so I'm curious why JDBC is your technology of choice. On Oct 20, 2015 1:12 AM, "Mihkel Nõges" wrote: > Yes I saw that issue too and set myself as watcher of this when it was >

Re: [Artemis] Master fails to start up after failback

Yes I saw that issue too and set myself as watcher of this when it was created. I did not think it could be exactly the same as it is described to present itself only in narrow timing related conditions. My case seems to be much more broad and basic. Seems like nobody actually tried to set this up

Re: [Artemis] Master fails to start up after failback

Looks related to me: https://issues.apache.org/jira/browse/ARTEMIS-256 On Mon, Oct 19, 2015 at 4:04 AM, Mihkel Nõges wrote: > Basic flow of getting unresponsive failback cluster: > Have machine with Ubuntu 14.04.3 > >1. Install libaio1, Java 1.8.0_60, maven 3.3.3, download and extract >

Re: [Artemis] Master fails to start up after failback

Basic flow of getting unresponsive failback cluster: Have machine with Ubuntu 14.04.3 1. Install libaio1, Java 1.8.0_60, maven 3.3.3, download and extract apache-artemis-1.1.0-bin in /opt

Re: [Artemis] Master fails to start up after failback

Hi Clebert, I do not have other code to share with you but the example code in Artemis 1.1.0 binary deployment package. I'm running org.apache.activemq.artemis.jms.example.ReplicatedFailbackExample And only commented out the serverStart and killServer calls which I am doing manually. I do not th

Re: [Artemis] Master fails to start up after failback

Im not on my computer now but it sounds like you are doing a fail back immediately after failed over. It takes some time (seconds) to the server to activate on the backup. Later the server will need to copy the data back before it can be activated in fail back mode. It sounds the live is not

Re: [Artemis] Master fails to start up after failback

Hello again! I would be very grateful If someone could answer my questions. We need the high availability to work to use the broker in production. When I run the replicated-failback example in one machine (broker3) it succeeds. It fails when I run the same test - exactly the same servers with sl

Re: [Artemis] Master fails to start up after failback

Hi Clebert, No test, just doing it on command line with standalone servers. I'm using 1.1.0 installed with apt-get, not the snapshot. I'm wondering what should be the suggested procedure for admins to do changes to HA cluster of 2 or 3 nodes of Artemis. If one of the nodes is master by configurat

Re: [Artemis] Master fails to start up after failback

The master needs to copy its data from the backup back to live before it's activated. Do you have a test replicating this? Did you try the snapshot build? On Tue, Oct 13, 2015 at 11:58 AM, Mihkel Nõges wrote: > Hi, > > I configured replicating HA master-slave of Artemis 1.1.0 instances on > Ubu

[Artemis] Master fails to start up after failback

Hi, I configured replicating HA master-slave of Artemis 1.1.0 instances on Ubuntu 14.04.3. When I kill master the slave takes over as expected and starts serving as new master. When I then start the old master, it fails with the following errors in the log: 16:35:46,476 ERROR [org.apache.activem