Troubleshooting
Default Components#
Cluster Manager utilizes the following components:
-
Gluu Server: free open source software package for identity and access management
-
Redis-Server: a value key-store known for it's high performance, installed outside the chroot on all servers. This is used a cache storage mechanism for tokens, session data and general caching of all the oxAuth and oxTrust/Identity services. The configuration file is located on the servers with Gluu at /etc/redis/redis.conf or /etc/redis.conf
-
Stunnel: used to protect communications between oxAuth, oxTrust and Redis with a proxy bounce through Twemproxy. The configuration file is located at
/etc/stunnel/stunnel.conf
on all servers. It runs on port 8888 of the NGINX/Proxy server and 7777 on the Gluu servers. For security Redis runs on localhost. Stunnel faciliates SSL communication over the Internet for Redis, which doesn't come default with encrypted traffic -
Twemproxy: (Optional) used for cache failover, round-robin proxying and caching performance with Redis. The configuration file for this program can be found on the proxy server in
/etc/nutcracker/nutcracker.yml
. It runs on localhost port 2222 of the NGINX/Proxy server. Twemproxy enables high availability by automatically detecting Redis server failure and redirecting traffic to other working instances. Twemproxy will not reintroduce failed servers. Restarting Twemproxy can be performed manually, or a script can be written to automate the task of resetting the "down" flag of the failed server. A proxy server is required for HA, unless you use redis cluster. Additional documentation about that can be found here -
NGINX: (Optional) used to proxy communication between Gluu instances. The configuration file is located on the proxy server (if used) at
/etc/nginx/nginx.conf
. If you are using an external HTTP load-balancer, this is not a necessary component. Session stickiness will need to be handled for all paths with the exception of/oxAuth
Logging for Errors and Troubleshooting#
Cluster Manager displays logs in the GUI about what's happening on the system it's interacting with.
There is also additional logging information in the $HOME/.clustermgr/logs
directory of the user who installed Cluster Manager.
If you have any other issues or concerns, please open a ticket at support.gluu.org.