Microsoft-Active-Sync crashing after Exchange database repair
My Exchange 2003 server had a corrupted database. I ran eseutil /p and fixed it but now my ActiveSync crashes whenever any thing tries to use it. Event ID 1011, source W3SVC. Description: A process serving application pool 'ExchangeActiveSyncPool' suffered a fatal communication error with the World Wide Web Publishing Service. The process id was '3584'. The data field contains the error number.
I recreated all the virtual directories already. OWA works fine.
ExchangeMicrosoft Server OSMicrosoft Legacy OS
Last Comment
Doug Penneman
8/22/2022 - Mon
Alan Hardisty
After you run eseutil /p, you should also run eseutil /d and isinteg -s servername -fix -test alltests
Did you?
Alan
Doug Penneman
ASKER
No, I'll do that tonight. Eseutil /p took 11 hours with a 25gb database.
Alan Hardisty
Yep - it isn't fast! About 4-6gb per hour depending on hardware, so 11 hours seems quite slow.
To resolve this problem, install Update Rollup 2 for Microsoft Exchange Server 2007 Service Pack 1. For more information about Update Rollup 2 for Exchange Server 2007 Service Pack
The above response doesn't make sense, since this post is for Exchange 2003 and not 2007.
@DougPenneman:
Please read the following hot fix.
The worker process for Exchange ActiveSync crashes when end-users try to synchronize their third-party devices with an Exchange Server 2003 server. http://support.microsoft.com/kb/957191
25gb.edb + 15gb.stm / 4gb/hr = 10 hrs. (is this how you calculate?)
Hardware is a Xeon E5-2620, 16gb RAM ECC. RAID 1 setup.
By the way, this Exchange server is a virtual machine on this host. That may be a discussion for another topic. Seems sluggish.
Eseutil /D took 7 hours.
Yep - that how you do it. It depends on the hardware, but with only 4Gb ram max, it isn't fast.
7 hours is better.
Isinteg is pretty quick and you should run it at least twice.
Doug Penneman
ASKER
My original problem still exist. Activesync becomes disabled and I also discovered my RPC over HTTP is not working either. It stops at connecting to port 6001. I've checked all the settings necessary for this to work but... could this be registry security messed up. After the original crash, the server did say Rebuilding Active Directory indices.
Alan Hardisty
Okay - is this SBS 2003 or Windows and Exchange 2003?
If SBS - please re-run the connect to the internet wizard. If not, please run through my article and check your settings:
After eseutil /d and isinteg fix (errors fixed), the MAS is working and the RPC over HTTP is working but I still get errors but not so critical now.
Here is the popup error I get at boot time after login:
IIS Worker Process encountered a problem and needed to close.
Details: szAppaname: w3wp.exe szModname: msvcrt.dll
Connectivity tester results for Outlook Anywhere:
Attempting to ping RPC endpoint 6001 (Exchange Information Store) on server server1.firstrestore.com.
The attempt to ping the endpoint failed.
The RPC_S_SERVER_UNAVAILABLE error (0x6ba) was thrown by the RPC Runtime process.
I've setup RPC over HTTP many times so I know all the steps and have verified they are in place. I'll try the SP2 reinstall later, I tried and it gave me the IMF is older and needs to be uninstalled error. one thing after another.
Alan Hardisty
Use the rpcnofrontend tool in the following link to check and modify your registry keys accordingly:
Not sure you need to reboot after removing IMF - just re-install SP2 afterwards.
Doug Penneman
ASKER
My emails have stopped being delivered for some reason now. I've restarted all the Exchange services. They're just sitting in the queue. No Events in event log. Crap.
Installed SP2, Rebooted server, there is still 105 emails in the local delivery queue. I've sent a couple of test and they get delivered but some are still adding to the queue. What do I do to fix this?
Alan Hardisty
Who are the emails destined for? Local email addresses that you recognise?
Yes they are all good local addresses. A few spams in there, but that's normal.
I would like some advice at this point.
Do I need to run isinteg again?
Should I create a "dial tone" database and move my current priv1 to it?
Should I create a new mailbox store and move the mailboxes to it?
Do I need to begin to think about recreating the server from scratch? ugh.
This was a new server box, setup as a virtual host. This Exchange server is a virtual machine on the box. One of the RAM modules was bad (most likely, STOP 7A errors) and went crazy and crashed the server a few times before I could replace the RAM.
This could be more corruption than just the mailbox store?
It is Exchange Enterprise. I've never created a new database and moved the mailboxes on the same server but it looks easy right?
I think I'm going to have to call Microsoft. It just won't straighten up.
Alan Hardisty
Not sure you will get any help from Microsoft with the server being virtual.
Doug Penneman
ASKER
Solution was ESEutil /P, ESEutil /D, isinteg x 3, create new store and move mailboxes that were not receiving mail. I did lose two days worth of email from last week but was able to recover them from the ost file.
Did you?
Alan