"Requesting data from Microsoft Exchange"

[MichealC4]

This message is appearing more and more for people on our install when people have their Outlook open and minimized. I don't use our Exchange server, so I haven't experienced it myself. I have searched around google, but so far everything hasn't been answered.

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[Zelandakh]

You have a bottleneck. Could be processor, RAM, disk etc.

Run perfmon counters on those and see what is running high.

 

[WhoKilledKenny]

The problem may have to do with disk I/O. Run performance monitor and look at physical disk where the database is, the counter is "Average disk second read and write" - The Average value should be below 10ms, for spikes, for best performance. If you are greater than 10ms spikes there is where your issue lies. This occrance is common with Exchange servers connected to a misconfigured SAN.

 

[MichealC4]

I took a look at memory usage and saw that store.exe was using close to 700,000KB of memory. Is this normal for store.exe? I know the Information Store should be pretty heavy, but I wouldn't think that heavy ... Then again, I'm not the expert in Exchange. perfmon is my next stop as soon as I can get away from everything else.

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[MichealC4]

Okay, during the time that I watched perfmon, here's what I got:

Avg. disk sec read: Remained close to 0 the entire time.
Avg. disk sec write: Highest spike was at .043 and average is .016

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[Ehobbs]

Is there any chance you are running a backup of some type when ths messages pops up? Lots of database/log backups will occupy the server enough to keep it from responding quickly.

Ed

 

[nickpark]

does it's appearance coincide with delivery of a message to them?

 

[MichealC4]

Ehobbs: No, I have an automated backup running every Sunday at 1AM.

nickpark: No, because one person I saw got it when their Outlook was just sitting there.

New (?) development, people are getting delivery delayed messages. I can't find anything that appears to be taking a lot of resources (other than what I've noted already). Right now, it is both a frontend and backend. But as near as I can see, it doesn't get that much usage on the frontend. We have a 200 user license, and something like 40 users on the server right now. Surely this isn't normal?

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[WANguy2k]

Store.exe will take as much RAM as it can, that's the way it's designed so that's probably not your problem. Are you sure there's nothing else running on the machine slowing it down? I had this problem when I had Brightmail (spam filter) installed on my Exchange server. I put Brightmail on another machine and it stopped.

Check the SMTP queues and see how much ougoing mail you have pending. If you are having an internet,DNS, or other connectivity problem (check your switch for errors on the port the Exchange server is plugged into), the server could be constantly retrying mail delivery slowing it down.

Also, check your virus software. Most contain documentation telling you what processes and directories to exclude from scanning or monitoring. If it's not set, it can really slow you down (see here:

http://www.msexchange.org/articles/Properly-Configure-file-system-antivirus-Exchange.html)

 

[xmsre]

Okay, during the time that I watched perfmon, here's what I got:

Avg. disk sec read: Remained close to 0 the entire time.
Avg. disk sec write: Highest spike was at .043 and average is .016 "

You didn't say which drive. If this is the disk with your logs on it, 43 ms is not a good time. I think the next step would be to look at the database - log record stalls/sec counter and see if log stalls coincide with your peak slow disk times.

 

[nickpark]

I assume no event log errors.
Maybe it's the network... while the message appears... does the server respond to ping?

You're the exchange administrator and don't use exchange!??

 

[MichealC4]

xmsre: RAID 5 3 disk, and 1 primary and 1 extended partition.

nick: Not that I have seen, no.

Funny thing about that, I was kinda handed the install and told set it up without having much knowledge of Exchange. I'm the guy that handles the security, not email. Email is someone else. Don't ask me why I'm the exchange admin, I just am. :/ And I don't use it because I don't like it. Besides, Thunderbird doesn't have support for Exchange.

At any rate, I'll check more in to that. I have another more pressing matter I'm currently working with Microsoft on. Keep the ideas coming though.

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[Zelandakh]

How many users on that box and what else is it doing? I'd say you've got disk IO issues.

 

[MichealC4]

88 users. All it is doing is Outlook Web Access and Exchange. My fault for forgetting (my excuse is that I have a number of unrelated tasks on my lap right now) to mention it, we recently moved from McAfee to Symantec on Exchange. Some users reported seeing it before the move (why they actually told us about it after the move I will never know ...), but most reported seeing problems after the move with no problems before.

Now, before anybody says it, yes, we should have tested first. The problem is, we were told by higher-ups to move and given a 4 day timeframe to move roughly 650 clients and servers from McAfee to Symantec. No time to test whatsoever. Personally, I'm not entirely sure this is the whole problem. Perhaps a lot, but I'm not sure it is the entire problem.

----------------------------
"Will work for bandwidth" - Thinkgeek T-shirt

 

[WhoKilledKenny]

Me again...
Wow, lots of discussion regarding your issue. You said you are working on another issue with Microsoft; I would open up another case with them regarding this issue. Sorry for the late reply but from you Perfmons looks like you are having an I/O issue.
You should be running your transaction logs on RAID 1 and your Databases on a separate array either RAID 5 or RAID 1+0. The logs and the database should be on separate spindles not only for performance but for disk crash and database recovery.
Sounds like you have inherited a bunch of work, my hats off to you and good luck...

 

[inspiron]

Are you running Full Text Indexing? If so, I would recommend switching it off. Sorted our problem immediately.