2

I've got pretty detailed proposals in front of management for either a full new on-prem build-out or moving to 0365. I'm just waiting on a decision, then purchasing/licensing, and implementation will still take some time.

We're an SMB running on-prem Exchange 2007 on an equally old Dell Poweredge server. Its Win 2008 R2, 8 GB ram, and has 2 E5405's @ 2.00Ghz. Everything is run out of this one box.

Thing have been getting progressively terrible over the last year. I've only been here about a year and trying to fight against the bad practices approach taken when it comes to some of these super critical servers. I thought we'd move on a 2016 on-prem or 0365 solution months and months ago and here we still sit pondering.

We have about 154 mailboxes for maybe 105-115 active users.

I have 4 active data stores/databases and 1 Public folder database. They are 264GB, 223GB, 109GB , 20GB and 12GB. I didn't think we could even have a store over 250GB, so someone must have fudged that limit at some point. I don't think its even appropriate for any of these to be over 100GB, based on how this server runs.

So a lot of people are complaining about pauses and lag in Outlook 2013. The folks who use the Salesforce for Outlook plugin have it about 1,000 times worse. I've been troubleshooting that from the SFO plugin side for months now, and since nothing has developed in doing a forklift upgrade on our mail servers I am now looking at it more from the Exchange end. I'm pretty sure Exchange 2007 on this server is the root cause of all of my woes.

Can anyone confirm that for me? More-over, what would be the best thing to break up these data stores? Is 100GB the right thing to aim for? Should I go for 50GB? Can we have a lot more stores with the specs we have on the server and actually see any sort of performance gain? Or is that a pipe dream? What else could buy me some temporary relief? Is more RAM worth it?

Additionally... I'd also like to know if anyone can confirm my suspicion that this Exchange server being so bloated is causing lots of problems down the chain.

We DO NOT run Outlook in cached exchange mode. There is some other utility we run that caused an issue with cached mode (We have an ancient ERP system running on SCO-UNIX and there's a little application we use that allows folks to "print" from the ERP to a PDF file that gets dropped as a e-mail attachment through MAPI.) I may have an alternate option for that process altogether, but its not a 'better' solution, just a side solution that doesn't put so many demands on Outlook. I'd need to explain that part of what is killing Outlook and the Salesforce plugin is that we're not running in cached exchange mode.

Our folks also seem hyper obsessed with creating folders. TONS of folders. Folders for everything. For customers, then broken down by events the customers has, and then broken down by orders... its crazy.

So ANY suggestions would be greatly appreciated. WHAT can I do to make this situation better while we SLOWLY get something new and better in place??

Sam K
  • 506
  • 5
  • 20
  • I can't help but think that no optimisation at all is the right answer. As more (and more senior) people complain the higher up the business priority stack the migration goes. – roaima Jul 06 '18 at 18:11

1 Answers1

1

The answer here was to throw 32GB of RAM in it. Still waiting on finalizing our 0365 migration but at least things are functional!

Sam K
  • 506
  • 5
  • 20