Workflow and ServiceDesk Community

Dear All,

I have a server where the first time a page, such as the Advanced Technician Submit Page, is "rendered", then it takes 5 to 10 seconds to show, then, once cached, the same page arrives instantly (sub-second). However, after a few minutes, the cache seems to be dropped and subsequent page loads take ages.

This is resulting in the system looking really show, if the user is not using the system very often, then the page loads always look like they take 5 to 10 seconds.

I think that this is a configuration issue for SD or IIS, so does anyone know what I can do to stop the cache from getting emptied?

Kindest regards,

QuietLeni

One place to check would be the IIS app pool recycle settings (regular time interval). I think by default the set time is like 20 minutes. There are likely other places, but I can't think of them off the top of my head.

Would increasing this regular time interval help or cause other issues?

It would help. That's that interval that determines if the pool should be refreshed if it hasn't received a page request in x number of minutes. If you have it set to 30, then the pool will empty if no one requests a page in 30 minutes. This means that unless someone is requesting a page at least once every 30 minutes, it will have to be recached.

Thanks, Michael.

If I set it to be 4 hours, then will this mean that more memory will be consumed in IIS?

Yes and no. Yes, The memory will be consumed for 4 hours before it is released if no one is accessing that page. No, that shouldn't be a problem as you were expecting people to hit that page and use the memory anyway so you already provisioned the server appropriately. ;)

For what it's worth (I am not an IIS expert!), each IIS app pool doesn't seem to ever use more than 2GB anyway, so it shouldn't be a big deal as they do tend to have a ceiling. On my SD install, I've got several app pools (one for incident projects, one for change, one for the process manager, etc.) and it seems to make better use of RAM that way, rather than one app pool that's hits it's 2GB limit and would need to cycling things in and out in order to stay under the limit. I'm not sure this is exactly how Symantec suggests doing things, but I am happy enough with the performance compared to when I had it in one pool.