A Forum reader recently asked this question:
“Essentially, what is the benefit of running agents in protected mode (what does that actually do?)? And if there’s a benefit, why wouldn’t you want to run entire system’s agents in protected mode all the time ?
And here are the responses from Jim Michael, Tim Heywood, and Scott Graham …
The benefit is that if an agent running in protected memory crashes, it won’t bring the other agents or applications down with it. What does it do? It literally runs in a new “memory space” in which it thinks it’s the only application running. This is a function of the CPU and NetWare.
You wouldn’t want to run entire system’s agents in protected mode all the time, because (paradoxically) in my experience running things in NetWare protected memory can cause more instability that it solves …
It isolates the function from the others running on the system a little more than without, it also provides a restart mechanism, so that if it would have abended the server in normal mode, it squishes the memory space and then reloads the software.
It is also useful if you are having issues with software not unloading. While it still will not unload from the protected memory space, you can also kill the memory space itself.
Also, if you make the agents restartable it sure makes your life simpler. For example, instead of getting a call about an agent being down, it fixes itself. Web Accesses Doc viewer used to be a major paina without protect/restart.
I run everything – Groupwise GWIA, MTA, POA’s, Web Access – in protected mode with restart enabled. The GWIA could have some issues because of this, such as lost mail, but I haven’t seen this problem.