Hi Thiago I have a "HA" setup currently running within Azure We found to many timeout issues with using a PAAS solution for MYSQL so instead are running a simpler solution at the moment.
We have two servers, one is the Master and one is the slave. We have a load balancer (traffic manager for setting priority) Which selects server number 1 (master) We are locking down the environment using oauth2 followed by LDAP and MYSQL. Server 1 (Master) has a RW database and syncs data to Server 2 (Slave) Guacamole on server 2 only have Read Only access to its database. If server 1 goes off line, server 2 can and will continue to allow connectivity however it will not allow creation of new users or connections, nor will it log who is logged on etc It is not a perfect but it does allow server 1 to be patched and/or go offline. From: Thiago dos Santos Nunes [mailto:[email protected]] Sent: Friday, 21 July 2017 1:21 AM To: [email protected] Subject: Implement HA on Guacamole Server Hi everyone, Pax! I need a lot of help. We have a guacamole setting with approximately 100-200 simultaneous connections. And we're investigating the option of creating client high availability using Hazelcast or memcached (Nick's tip). Has anyone tried this? Could you share how it was? Because I've never worked with it. Another essential thing is the server side. We need to implement some High Availability schema for the server. And it would have to be something without downtime if possible. I had already created a ticket for the HA issue on the guacamole server, in case there is a need to change the code. This would be an exceptional feature for medium to large environments. Https://issues.apache.org/jira/browse/GUACAMOLE-283 Please help me in this, because the environment has fallen sometimes and the users get very frustrated .... And customers wanting to paralyze services for this. Stay with GOD! Aude et Effice! Thiago.
