Good Morning,
I have a question regarding mail routing between Source and Target domains.
I'd like to know how the mail flow works after additional addresses (source.local and target.local) are set up.
Does Qmm proxy mail between domains or the flow it's done between the Exchange environments with no qmmex working in the middle of the mail routing?
Basically is it SOURCE <--> TARGET or SOURCE <--QMMEX--> TARGET
Thank you
Slav
QMM does not handle the routing - you have to setup your Connectors in Exchange to support the routing of these namespaces.
I have needed connectors in place. I need to understand the mail flow since I have the source Exchange located a different geo location than my target Exchange. The existing mail flow is very slow. I was wondering is setting up an additional target server and creating a new dag (target Exchange) in the same city the the source is could fix my slowness issues.
If the mail routing is not done thru my qmm that means the above idea could work.
That's really hard to say.
Where is the slowness originating - slow Exchange servers or connectivity or maybe both?
Do you have a new network link between the sites that your "QMM" connectors can use?
I don't see how adding a new server to the mix will help unless your existing infrastructure is overworked on both sides right now?
The Exchange environments are doing ok performance wise. The problem is that mailbox syncing in QMM is very slow. Other than that, there is one more problem - the average mailbox size is 50GB with over 100k objects on average. The largest mailbox to sync is 100GB and 300k objects. There is nothing I can do to shrink those mailboxes which have to go to the Target as is (no size changed).
Also the idea was to add an Exchange server for the Target domain in the same data-center where the Source Exchange is hosted from. That's to avoid the slowness
Add more Agent host servers, with more collections , and just a few MBX per collection. Put the sleep time to a minimum amount 5 mins tops.
I did that. Still slow. I have 10 agents. So 1 agent can do 5 users but as I said the problem is no mimecast in place or any other archiving solution.
All into MBX without archiving prior? The Agent Host server specs? Is the local AV scanning the Agents?, if so exclude them
Also you have Riverbed in place, between both places? We had a problem with them caused issues with the Sync, to fix we made the traffic go through unimpeded. (Src > Trg)
There is no archiving approved, AV is there but there are exclusions in place. AV specs: win 2012 r2, RAM 8GB, 4 cores, cpu utilization ~30%
win 2012 r2, RAM 8GB, 4 cores, cpu utilization -30% = Agent Host Server?
Also find out concerning the Riverbeds
