We have and manage several installations of ASG, all in HA with active/standby. Servers we use for each pair in HA are always identical in configuration, and all of well know brands (HP Proliant, IBM X Series, Supermicro, etc.). All servers are installed and running within a tier IV class datacenter and networking gear around the ASG are either Cisco, 3Com or HP.
The licenses we have are either Network Security, Web Security of full.
In all cases hardware performance in average is normal, no or minor swap, hard drives are either solid state or in RAID mode with RAID controller and cache, 4 to 8 network ports (Intel of Broadcomm), some of the servers have redundant power supplies or are connected to an automatic power transfer switch, etc.
Current ASG version is 8.301
We noticed that after few weeks, 2 to 3 in average, the server that is acting as master starts to do strange things (e.g. slower performance, slower the network, VPN disruptions, etc.). As of now, the way we found puts everything back to normal is simply to reboot the server that is acting as active, let the standby takes the role of active and the symptom repeats but now on the other server and a few weeks later the cycle repeats again.
This behavior sounds strange, it is clearly independent of the hardware (server, network gear, etc.).
In the ASG logs we could not find any information or errors.
I would like to know if there is previous history, record, workaround or log about this and if so, what are the plans of Astaro/Sophos to fix this.
Thank you in advance.
The licenses we have are either Network Security, Web Security of full.
In all cases hardware performance in average is normal, no or minor swap, hard drives are either solid state or in RAID mode with RAID controller and cache, 4 to 8 network ports (Intel of Broadcomm), some of the servers have redundant power supplies or are connected to an automatic power transfer switch, etc.
Current ASG version is 8.301
We noticed that after few weeks, 2 to 3 in average, the server that is acting as master starts to do strange things (e.g. slower performance, slower the network, VPN disruptions, etc.). As of now, the way we found puts everything back to normal is simply to reboot the server that is acting as active, let the standby takes the role of active and the symptom repeats but now on the other server and a few weeks later the cycle repeats again.
This behavior sounds strange, it is clearly independent of the hardware (server, network gear, etc.).
In the ASG logs we could not find any information or errors.
I would like to know if there is previous history, record, workaround or log about this and if so, what are the plans of Astaro/Sophos to fix this.
Thank you in advance.