So... You have three ActiveMQ nodes that are capable of failing over in the case of a problem with the current master broker. That's great. But your NFS server is a single point of failure, and if it goes down, all three ActiveMQ servers are useless. You might want to figure out how to make your NFS server be clustered/redundant, to ensure availability in the case of hardware failure.
Tim On Thu, Aug 15, 2019, 6:21 PM Arshkr <phu.ngu...@unisys.com> wrote: > Hi Tim, > > I think I got it works. Here is my note, in case any newbie like me have > some idea where to start from :-) > > Here is what I did for 3 ActiveMQ servers, and 1 NFS file server: > > - Set up an NFS share server on windows 2018, and create a shared folder > name "activemqdata". Instruction on how to setup: > > https://docs.microsoft.com/en-us/windows-server/storage/nfs/deploy-nfs#deploy-nfs-infrastructure > > - download ActiveMQ v5.x and copy to 3 Windows 2018 servers, ActiveMQ is > locate at C:\activemq on each server. > Instruction on how to setup: > https://wiki.eveoh.nl/pages/viewpage.action?pageId=14287030 > > - From each ActiveMQ server, map a network drive activemqdata\ as a local > drive and assigned a drive letter. In this case, I use B:\activemqdata > > - move a data\ directory in ActiveMQ installation to B:\activemq\data, and > deleted remaining "data\" on last two servers > > - modify activemq.xml in conf\ directory to use kahadb in > "B:\activemq\data" > > - Start ActiveMQ on each server, and everything works as expected. In my > case, I have 3 ActiveMQ servers and an NFS file server. > > > > -- > Sent from: > http://activemq.2283324.n4.nabble.com/ActiveMQ-User-f2341805.html >