CODEWITHSUNDEEP
activemq-classicmaster-slavekahadb
user1093643
user1093643

Reputation: 271

File locking not aquired in master-slave cluster

I´m trying to configure a clúster of ActiveMQ (v.5.12.1) instances in master-slave using a shared file system.

Those instances are running in different servers using this configuration

<persistenceAdapter>
   <kahaDB directory="/sharedDir/data/activemq/kahadb"/>
 </persistenceAdapter>

The problem is that both instances start as master because lock file mechanism does¡t seem to be working.

I´ve made a test with 2 instances in the same server, and with this configuration it works as expected (2nd instance can´t acquire file lock and waits as slave).

We are using NFSv4 and mount properties are like this:

x.x.x.52:/mnt/data on /mnt/nfs type nfs4 (rw,relatime,vers=4.0,rsize=8192,wsize=8192,namlen=255,hard,proto=tcp,port=0,timeo=14,retrans=2,sec=sys,clientaddr=x.x.x.167,local_lock=none,addr=x.x.x.52)
bindfs on /sharedDir type fuse.bindfs (rw,relatime,user_id=0,group_id=0,default_permissions,allow_other)

Thank you

Upvotes: 0

Views: 530

Answers (0)

Related Questions