I think there may be some confusion about how things actually work.
Multiple active brokers should _never_ be using the same journal. Each
broker in a cluster should have its _own_ journal. The only time brokers
should share the same physical journal is when they are configured for high
availabilit
Justin,
When I say "shared File System" I mean all nodes of the cluser, which are now
all active simultaneously having access to same journal.
On Friday, August 9, 2024 at 10:38:08 PM GMT+10, Justin Bertram
wrote:
Why do you need a shared file system if you're not actually sharing fil
Why do you need a shared file system if you're not actually sharing files?
Are you referring to EFS as a "shared" file system simply because it is a
_network_ file system? Please clarify.
Generally speaking, ActiveMQ Artemis should work fine with EFS.
Justin
On Fri, Aug 9, 2024 at 6:29 AM Vimal
Thanks Jan. But in my case its a symmatric cluster i.e no aactive-passive but
all active active
On Friday, August 9, 2024 at 08:56:37 PM GMT+10, Jan Šmucr
wrote:
Eh, sorry, bad link.
Here:
https://activemq.apache.org/components/artemis/documentation/latest/ha.html#shared-store
Jan
-
Eh, sorry, bad link.
Here:
https://activemq.apache.org/components/artemis/documentation/latest/ha.html#shared-store
Jan
-Original Message-
From: Jan Šmucr
Sent: pátek 9. srpna 2024 12:55
To: users@activemq.apache.org
Subject: RE: Need architecture advice
Hi Vimal,
yes, it can, since E
Hi Vimal,
yes, it can, since EFS is only a bit limited version of NFS v4.1. You just need
to be aware of potential throughput and latency limits. It's definitely a good
thing to do some benchmarking first before you fully commit to this solution.
See:
https://activemq.apache.org/components/arte
Can artemis symmatric cluster be implimented using shared file system (EFS) on
AWS?
