Thanks for the responses. I've made sure I'm running on the latest version (7.0.6).
I've made sure port 5701 is open thru the firewall on both servers, using telnet to verify from each. I've set up the following config on both servers: cas.ticket.registry.hazelcast.cluster.core.instance-name=login-test cas.ticket.registry.hazelcast.cluster.network.members=10.20.2.106,10.20.2.105 cas.ticket.registry.hazelcast.cluster.network.port=5701 cas.ticket.registry.hazelcast.cluster.network.port-auto-increment=true cas.ticket.registry.hazelcast.cluster.network.tcpip-enabled=true cas.ticket.registry.hazelcast.cluster.discovery.multicast.enabled=false cas.ticket.registry.hazelcast.cluster.core.replicated=true cas.ticket.registry.hazelcast.cluster.core.async-backup-count=0 One of the two servers never seems to make it all the way "up" as I don't ever see it go to "Ready" state or show the periodic JsonServiceRegistry messages, [org.apereo.cas.services.mgmt.AbstractServicesManager] - <Loaded [40] service(s) from [JsonServiceRegistry].> If I check both servers today, I can see the following error message related to hazelcast: node1: ERROR [com.hazelcast.internal.cluster.impl.ClusterHeartbeatManager] - <[10.20.2.106]:5701 [dev] [5.3.6] Connecting to self! [localhost]:5701> node2: ERROR [com.hazelcast.internal.cluster.impl.operations.MasterResponseOp] - <[10.20.2.105]:5701 [dev] [5.3.6] Connecting to self! [localhost]:5701> I do run the tomcat behind an on-board apache httpd proxy, but it is the same configuration I ran under 6.6 where hazelcast worked. Not sure what to check next. Phil On Friday, August 2, 2024 at 12:47:03 PM UTC-5 Ray Bon wrote: Erik, Increase tomcat memory; to 16 or more Gb. Ray On Thu, 2024-08-01 at 13:34 -0500, Erik Mallory wrote: You don't often get email from erik.m...@gmail.com. Learn why this is important <https://aka.ms/LearnAboutSenderIdentification> Same for 7.0.4 in production our test and dev environments work but not prod. We have some style changes (colors) we use to identify the different environments and I use the war overlay to rebuild the war and copy it to the environment. dev works, I finished work on the 7.0 branch a few months back and upgraded test two months ago, and upgraded both environments to 7.0.4. The move to prod earlier this week did not go as well. We experienced the same thing you are. with 7.0.6. Moments ago I upgraded our dev system to 7.0.6 to see if I could replicate your problem and I could not. Hazelcast is up and running and cas works properly. So what's different? our prod cas gets HAMMERED. test and dev not so much. Through Prometheus, we noticed that the Old_Gen Memory for our production system was growing under load and eventually would tank the cas service when the jvm ran out of memory. I'm not sure where the problem is, but I don't think it's your config. On Thu, Aug 1, 2024 at 11:23 AM Phil Hale <phal...@gmail.com> wrote: Howdy folks, I'm having some trouble with Hazelcast after upgrading from CAS 6.6 to 7.0. I have CAS servers set up in an HA configuration and have been using Hazelcast for the Ticket Registry for many versions of CAS. After upgrading to 7.0, the basic configuration I've used has stopped working. Generally I just had to include the following two cas.properties settings: node1: cas.ticket.registry.hazelcast.cluster.core.instance-name=10.20.2.106 cas.ticket.registry.hazelcast.cluster.network.members=10.20.2.106,10.20.2.105 node2: cas.ticket.registry.hazelcast.cluster.core.instance-name=10.20.2.105 cas.ticket.registry.hazelcast.cluster.network.members=10.20.2.106,10.20.2.105 But the two hosts no longer seem to communicate properly. I'm on CAS version 7.0.6. I've reviewed the Required configuration settings from https://apereo.github.io/cas/7.0.x/ticketing/Hazelcast-Ticket-Registry.html but I still can't seem to get it working. Anyone have a sample configuration for a 2 node setup that is working under 7.0 that they can send me? Thanks, Phil -- - Website: https://apereo.github.io/cas - List Guidelines: https://goo.gl/1VRrw7 - Contributions: https://goo.gl/mh7qDG --- You received this message because you are subscribed to the Google Groups "CAS Community" group. To unsubscribe from this group and stop receiving emails from it, send an email tocas-user+u...@apereo.org. To view this discussion on the web visit https://groups.google.com/a/apereo.org/d/msgid/cas-user/ea7920bc-854d-47e3-9280-6bb9e4c13d39n%40apereo.org <https://groups.google.com/a/apereo.org/d/msgid/cas-user/ea7920bc-854d-47e3-9280-6bb9e4c13d39n%40apereo.org?utm_medium=email&utm_source=footer> . -- Erik Mallory ------------------------ "A happy man's paradise is his own good nature." - Edward Abbey -- - Website: https://apereo.github.io/cas - List Guidelines: https://goo.gl/1VRrw7 - Contributions: https://goo.gl/mh7qDG --- You received this message because you are subscribed to the Google Groups "CAS Community" group. To unsubscribe from this group and stop receiving emails from it, send an email to cas-user+unsubscr...@apereo.org. To view this discussion on the web visit https://groups.google.com/a/apereo.org/d/msgid/cas-user/4a3ef95b-6aec-4792-b0cf-6a26c059ee47n%40apereo.org.
