Unable to join the existing portal

649
1
01-07-2020 08:38 PM
FrankXu
New Contributor

We are trying to setup HA for ArcGIS Portal 10.7 on Linux (RHEL7.7)

I have followed the steps in Configure a highly available portal—Portal for ArcGIS (10.7 and 10.7.1) | ArcGIS Enterprise.

I have the primary portal up running, and setup a content directory on NFS on a separate server, and have the primary portal using the content directory on the shared NFS server. And opened ports 5701-5703, 7005, 7099, 7654, 7120, and 7220 which are required for HA on both servers hosting the portal.

However, After I installed the second portal, and when I attempt to join the existing primary portal , I got an error:  

There was an error when attempting to join the existing portal.
Unable to configure local machine in standby mode for high availability.

When I check the logs, it shows it tries to connect to the primary server on port 5702 & 5703, but getting connection refused error, and they are then added to the blacklist by Hazelcast cluster? What does that mean? And then the log shows the below:

WARNING [hz._hzInstance_1_arcgis_admin.priority-generic-operation.thread-0] com.hazelcast.cluster.null [IP-xxx-xx-xx-xxx.AP-SOUTHEAST-2.COMPUTE.INTERNAL]:5701 [arcgis_admin] [3.11] Node could not join cluster at node: [IP-xxx-xx-xx-xxx.AP-SOUTHEAST-2.COMPUTE.INTERNAL]:5701 Cause: the target cluster has a different group-name

Where is the group-name defined? Anyone has any ideas and suggestions? Thanks in advance.

0 Kudos
1 Reply
DanielBrumm1
Occasional Contributor II

Were you ever able to resolve this issue? I have the same problem right now

Daniel Brumm
GIS Nerd
0 Kudos