I posted this into the S2K8 forum and decided this was more appropriate... Server2k8 64-bit with all current updates. Exchange 2k7 64-bit with all updates. IIS helper process is crashing every 5 minutes, or so. Here's the data for troubleshooting:System - Provider [ Name] Application Error - EventID 1000 [ Qualifiers] 0 Level 2 Task 100 Keywords 0x80000000000000 - TimeCreated [ SystemTime] 2008-06-17T01:29:40.000Z EventRecordID 51579 Channel Application Computer mailserver.mydomain.local Security - EventData w3wp.exe 7.0.6001.18000 47919ed8 davex.dll 8.1.278.0 47f36909 c0000005 00000000000d3be1 At the same time, I get this in the system log:A process serving application pool 'MSExchangeOWAAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '6972'. The data field contains the error number. I'm also getting this in my security logs at the same times... - System - Provider [ Name] Microsoft-Windows-Security-Auditing [ Guid] {54849625-5478-4994-a5ba-3e3b0328c30d} EventID 4625 Version 0 Level 0 Task 12544 Opcode 0 Keywords 0x8010000000000000 - TimeCreated [ SystemTime] 2008-06-17T18:18:44.706Z EventRecordID 2185746 Correlation - Execution [ ProcessID] 640 [ ThreadID] 4656 Channel Security Computer mailserver.mydomain.local Security - EventData SubjectUserSid S-1-5-18 SubjectUserName mailserver$ SubjectDomainName mydomain SubjectLogonId 0x3e7 TargetUserSid S-1-0-0 TargetUserName mailserver$ TargetDomainName Status 0xc000006d FailureReason %%2313 SubStatus 0xc0000064 LogonType 3 LogonProcessName Advapi AuthenticationPackageName Negotiate WorkstationName mailserver TransmittedServices - LmPackageName - KeyLength 0 ProcessId 0xf24 ProcessName E:\Program Files\Microsoft\Exchange Server\Bin\Microsoft.Exchange.Search.ExSearch.exe IpAddress - IpPort - This is causing instability in OWA. It's affecting rpc over http. This is generally making my life miserable. I've gone through and checked the IIS directory permissions as found on this post: http://forums.technet.microsoft.com/en-US/winserversetup/thread/864180f0-1720-49a1-a117-aa65609eced7 I'm also regularly seeing from https://mailserver/owa an error that says: 503 Service Unavailable Last, but not least... A process serving application pool 'MSExchangeOWAAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '5624'. The data field contains the error number. A process serving application pool 'MSExchangeOWAAppPool' terminated unexpectedly. The process id was '7220'. The process exit code was '0xff'. A process serving application pool 'MSExchangeOWAAppPool' failed to respond to a ping. The process id was '2296'. And these lead to... Application pool 'MSExchangeOWAAppPool' is being automatically disabled due to a series of failures in the process(es) serving that application pool. I really need to figure out why everything web-related to Exchange is behaving so poorly. Help?

You did not mention this, but it seems like you have pretty thoroughly checked everything, but are you running E2K7 SP1 with the latest roll-up (roll up 2)? I saw in the dump that it looks like the DLL in question is the DAVEX.DLL. That is not really used by a native E2K7 org (at least I don't think they are.) What URL do your users use to connect to E2K7? Do they use the /Exchange virtual directory at all?

Exchange SP1 with RollUp2 installed. In fact... with "other Microsoft Products" being updated, the only patches currently available that I haven't installed are Silverlight and 950049. As far as davex... We've got people using Entourage and Outlook 2k3 so I had to enable compatibility mode on the machine. the /Exchange virtual directory? I do believe that Entourage will use that when this finally gets up and running.

Hi, Event 4625 has been reported as a bug, it will be fixed later. It does not affect the performance of Exchange Server. Besides, event 1000 related to Davex.dllseems a performance issue. Let me explain how davex.dll works in Exchange Scenario. MicrosoftExchange uses Exprox.dlland Davex.dllto redirect requests that are submitted to virtual directories on Exchange2007, Exchange2003, and Exchange2000 computers to the correct URL. Specifically, Exprox.dll proxies client requests from the Client Access server to the Exchange2007 Mailbox server, the Exchange2003 back-end server, or the Exchange2000back-end server. The logic for handling legacy requests is provided by Davex.dll. Davex.dll handles Distributed Authoring and Versioning (DAV) requests, redirects Exchange2007 mailbox users to the /owa virtual directory, and displays the Exchange2003 and Exchange2000OutlookWebAccess experience. Exprox.dll only directs traffic to Davex.dll on a mailbox server. Davex.dll decides which server is the correct one to use. Be aware of the following when you work with Davex.dll: 1. Davex.dll responds to both DAV and OutlookWebAccess requests. If you are pointing your browser to a URL, such as https://mail.contoso.com/exchange, and Davex.dllresponds to it, your browseris treating the requestas an OutlookWebAccess request. 2. Davex.dll will redirect a request based on the internal (intranet) name of the server. This means that users on the Internet may receive a DNS error because the internal name of a server may not be the one that is exposed on the Internet. Then Davex.dll is frequently used by Exchange, when you try to access OWA. Wed like to recommend you to dispatch more ram for it. You may try to add Ram for your Exchange Server. More information for your reference: Outlook Web Access and Exchange 2007, Exchange 2003, and Exchange 2000 Coexistence http://technet.microsoft.com/en-us/library/bb885041(EXCHG.80).aspx The W3wp.exe process may crash in low memory situations in Exchange Server 2003 http://support.microsoft.com/kb/912586 DAVEX.DLL file is missing http://technet.microsoft.com/en-us/library/aa998539(EXCHG.80).aspx Managing Outlook Web Access Virtual Directories in Exchange 2007 http://technet.microsoft.com/en-us/library/aa998849(EXCHG.80).aspx Hope it helps. Xiu

Xiu - your post was extremely informative. thank you. I have a follow up question- I'm experiencing most of this scenarios described, but I'm running server 2008 x64, exchange 2007 sp1 with rollup 2. Is the w3wp.exe low memory crash also documented exch 2007? The KB you posted appears to indicate 2003 only... but I'm seeing this happening and would love to get access to a hotfix if availible. Thanks.

Hi, So far,I do not find any related hotfix. If you perfer,then you can use Adplus tool to catch the crash dump file to find more information on this issue. Debugging Tools for Windows(x64) http://www.microsoft.com/whdc/devtools/debugging/install64bit.mspx 286350 How to use ADPlus to troubleshoot "hangs" and "crashes"http://support.microsoft.com/default.aspx?scid=kb;EN-US;286350 Hope it helps. Xiu

We're having same issue, 7 hours on the phone with microsoft and no answers yet. Server '08 exchange '07 all updates installed. The iis worker process stops and then MSExchangeOWAAppPool stops. We thought it was a memory issue so we doubled the RAM in the server. No luck. When I restart the App Pool it just crashes again 10-20 minutes later.

I've now discovered that even though we doubled our RAM we're still using 95% constantly. I've got 4 GB of RAM in the system and it's always maxed out. I have about 200mb free memory and as soon as two people log into owa I go down to 0 and then I get my crash.

Well ***! I was just about to order more ram, hahaha. I figured throw more memory at a memory error

We went from 2 to 4 GB and that didn't help. Somehow have 1 gb cached and only 100 mb free, get two users logged into owa and I'm down to 0 within 2 minutes.

You didn't mention if you're running Ex07 or Ex03, but with Ex07, it will use all of the available memory to cache your message database. I'm running with 12GB of RAM, and it uses all of it. With memory being cheap now, it doesn't make sense to try to "get by" with low amounts. Most server-class motherboards support anywhere from 8GB to 32GB of RAM. Your problems, however, seem to be related to a memory leak or running out of system resources. If you're running Ex03, investigate the 3GB option and the following article (http://support.microsoft.com/kb/316739/). It also might be time for a new server & a new build of Exchange; thenyou can move mailboxes to something you can establish as a "known" working Exchange server.

My posted above and the subject reads that we are using exchange 07 and server 2008. Yes this is a memory leak. I also discovered something today that is interesting, may be coincidence but worth noting. Yesterday I disabled the Microsoft Exchange Monitoring service. This service callows applications to call the exchange diagnostic cmdlets. After a reboot OWA worked fine and then this morning one of my employees says their blackberry isn't working anymore. They sync their blackberry through the verizon website not BES. So I re-enabled the Exchange Monitoring service and their blackberry started it's sync again but as soon as it synced I noticed the that OWA errored out. A process serving application pool 'MSExchangeOWAAppPool' suffered a fatal communication error with the Windows Process Activation Service. The process id was '4392'. The data field contains the error number. I haven't had a chance to further test because he really needs his blackberry since he'll be out of the office today and tomorrow. As soon as he gets back I'm going to try some testing to really see if it's that service or his blackberry sync that is causing the crash.

We don't have any blackberries, but we do have winmo devices syncing with activesync, I also have random people using OWA in the field, or internally. I even have used it internally a few times when I'm not at my desk and need to get into my mailbox. I also have the outlook anywhere feature enabled, and have a remote user that uses outlook over http. I've seen the OWA issue from a user standpoint, and waiting about 30 seconds and refreshing will fix the UI. And sadly that's the instructions I've sent to my users. "if you're using OWA and something weird happens, wait 30 seconds and refresh" lol

It can't be a problem with activesync cause I can sync my treo and we don't get a trash, plus I have 3 other active sync devices on the network that sync with no problem. I'll do further testing on Monday and let you know. We have had a case open with microsoft for over a week but my lead tech is on vacation and I need to wait till he gets back and then he can follow up with microsoft. I know he was doing some debugging.

Hi, You can use ADplus tool to catch the dump file for further troubleshooting. Meanwhile, you can send the dump file to my mailbox xiu zhang. Thanks. Xiu

I appreciate your continued assistance Xiu - I'll schedule some downtime this weekend to provide the requested information and we'll see what we can come up with. Derek

Care to feed me a command line for ADPlus? The last technician I talked to spent almost 2 hours wasting my time trying to figure out just how to get the darn thing to run. The support ticket we have in with Microsoft has been pretty useless, thus far... just randomly guessing at a possible cause for this.PS - CHays... I'm back from vacation.

I'm having a similar problem with Server2008/E2k7SP1. I see events where the app pool configuration file cannot be accessed 5 times in a row and the pool thread terminates.

Hi, Below is the example when we use adplus to troubleshoot crash happen with microsoft.exchange.imap4.exe 1.Run Adplus.vbs on the Exchange server console 2.Open a command prompt and navigate to the debuggers folder (Default is c:\Program Files\Debugging Tools For Windows):cscript adplus.vbs -crash -pn [processname] -o c:\[processname]_DumpFileexample to create a dump file for the IMAP process: cscript adplus.vbs -crash -pn microsoft.exchange.imap4.exe -o c:\IMAP_DumpFile3.Click "OK" to the Adplus Warning stating "An _NT_SYMBOLS_PATH environment variable is not set..." 4.Review the Windows Script Host dialog box stating ADPlus is now running in CRASH mode" 5.Click OK after reading the dialog box.At this point, Adplus will attach a debugger to the process and wait for an Access Violation. Once an Access Violation occurs, the debugger will be invoked and a dump file will be created. Do NOT manually stop or terminate the process.The folder c:\IMAP_DumpFile will be automatically createdThe Store dump file will have a .dmp file extension and associated data files will be placed in the folder. Hope it helps. Xiu

Xiu Zhang - MSFT wrote: Hi, You can use ADplus tool to catch the dump file for further troubleshooting. Meanwhile, you can send the dump file to my mailbox xiu zhang. Thanks. XiuI am experiencing the hang issue as well on my end. (windows 2008 x64, exchange sp1 with rollup 2). I emailed over a bunch of crash dump files for the w3p process and the OWA pool crashing. Please let us know what you find!In case you don't get the email, the dump file is here.

Hi Thomas, I've got the dump file and will do some analysis. Besides,you can try the below steps to troubleshoot. 1. Reload davex.dll1. Stopped the default web site2. Renamed davex.dll to davex.dll.old(c:\Program Files\Microsoft\Exchange Server\bin)3. Started the default web site4. IIS restarted(iisreset /noforce) 2. Please check Anti-Virus Software to see whether it has killed w3wp.exe. Also, please confirm thatthe crash is caused by Davex.dll. Hope it helps. Xiu

I have the exact same problem. Server 2008 with Exchange 2007 SP1 and OWA application pools crashing. Have had a case in since 6/30/08. Have been escalated to level 2 support and then promptly blown off just in time for a holiday weekend. I have provided the dumps created by the x64 diags to several engineers - nothing. Let me confirm for poster - crash is caused by davex.dll. Can anybody please provide a solution?

We don't yet have AntiVirus on the server (other than Forefront for Exchange). So it's not the AntiVirus.killing off davex won't help us as we need Davex for Entourage clients. And EVERY time the OWA app pool crashes, the event log references davex.dll.We, too, have a ticket in with L2 support.

Hi, For better analyze the dump file, please try to gather the below information and post here. 1. Please check whether any3rd party applications running on Exchange Server. 2. Any specific apps other that OWA accessing OWA? 3. Only CAS role installed on the server? 4. How often do you get the error? 5. What hardware and what hardware management tools installed? (some have web services too.like HP) 6.Please check MSExchangeOWAAppPool properties,whether you have do some changes on it. Please post the information here. Besides, please install MPSRPT_Exchange.exe to gather more detailed system status and configuration information. Please send the bat file to xiu zhang. How to install the Exchange edition of the Microsoft Product Support Reporting Tool http://support.microsoft.com/?id=870640 Looking forward your reply. Thanks. Best regards, Xiu

Renaming davex.dll to .old worked for me. This was the only post that suggested this Thanks...

Glad to hear it works with you,Ryan.

The list of things I did to try to mitigate the issue is rather long and convoluted, and I'm not 100% certain at what point things settled down. I believe I was having some probems brought on by corruption in .net, the most noticeable improvements were after SP1 Rollup 3 was installed. As a whole, Exchange is much more stable after that. I still have lingering issues with activesync, but that's for a different thread. Thanks for all the help.

Hi, Since this thread has been opened for so long time. I would like to share more steps to troubleshoot this issue. 1.Try to uninstall CAS role and then re-install it. 1) Open up the control panel and select Add and Remove Programs. 2) Choose the Exchange 2007 program from the list and select Remove. 3) Uncheck the CAS checkbox to uninstall the CAS role and let the uninstall complete. 4) Open up the control panel and select Add and Remove Programs. 5) Choose the Exchange 2007 program from the list and select Change. 6) Select the CAS role to install and let the install complete. 7) Re-start Exchange Server. 2. If that do not help, then please check whether web.config tag <identity impersonate="true" /> (OWA, Auto discover virtual directory)has been set. If yes, then you can follow the below steps: 1. Modify the web.config file in the root of the virtual directory and change the following tag from: <identity impersonate="True" /> back to: <identity impersonate="false" />. 2. Restart your Exchange application pool. 3. Re-run your Powershell cmdlet. More information share with you: http://blogs.msdn.com/dgoldman/archive/2007/12/06/title-setting-the-web-config-identity-impersonate-true-will-cause-exchange-poweshell-cmdlets-to-fail-and-crash-the-msexchangeservicesapppool.aspx Hope it helps. Xiu

Xiu, will uninstalling/reinstalling the CAS role impact the configuration? Will there be any effect on this (ie my POP/IMAP config, etc...)? Thanks

When you say "Re-start Exchange Server" do you mean an actual server reboot or just restart the service? Xiu Zhang - MSFT wrote: 7) Re-start Exchange Server. Xiu

Hi StephenZF, Please backup the configuration that you has set manually, and then try to reinstall it. Hi chays1977, Please let the CAS Server re-start to apply the settings. Hope it helps. Xiu

Hey Xiu - I noticed the blog post referrs to IIS 6... does the same apply to IIS 7?

Hi Derek, The method should also can be donewithIIS 7.0. You may backup your web.config file in case any unexpected problem occurs. Xiu

Hi Xiu,when I go in Add/Remove Programs and modify the Exchange 2007 installation, the installed roles are grayed out and I can't uncheck the CAS role to uninstall it - any idea why?

Hi, In the Add/Remove windows, please find Exchange Server 2007,then click on "Remove",follow the setup wizard,and thencheck whether the CAS role can be un-ticked. Regards, -Xiu

OK - that works. I'll try it out and get back to you if I have any issues. Thanks

Don't do it. I did it on my server and it didn't help at all, ended up making matters worse. Now our free busy doesn't work so users can't check out of office. Microsoft techs have confirmed this is a bug I have a ticket open with microsoft. As new information comes available I'll let you know.

Thanks for the warning - I'll wait for feedback

I've got the same issue too. We're all still watching and waiting for the magical solution!

Having the same issue here. Its very frustrating, and is causing our Entourage users to have their email blocked quite often. They will send a few emails, then the next 2-3 will just sit in their outbox and blow error -17895. After a few minutes, when the app pool recovers, they can send emails for a while, then it will stop (either causing the app pool to crash, or the app pool is causing the issue). Getting many: Event Type:ErrorEvent Source:Application ErrorEvent Category:(100)Event ID:1000Date:8/7/2008Time:1:22:50 PMUser:N/AComputer:EXCHANGEDescription:Faulting application w3wp.exe, version 6.0.3790.3959, faulting module exprox.dll, version 8.1.291.0, fault address 0x0000000000026f7e. For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.Data:0000: 41 70 70 6c 69 63 61 74 Applicat0008: 69 6f 6e 20 46 61 69 6c ion Fail0010: 75 72 65 20 20 77 33 77 ure w3w0018: 70 2e 65 78 65 20 36 2e p.exe 6.0020: 30 2e 33 37 39 30 2e 33 0.3790.30028: 39 35 39 20 69 6e 20 65 959 in e0030: 78 70 72 6f 78 2e 64 6c xprox.dl0038: 6c 20 38 2e 31 2e 32 39 l 8.1.290040: 31 2e 30 20 61 74 20 6f 1.0 at o0048: 66 66 73 65 74 20 30 30 ffset 000050: 30 30 30 30 30 30 30 30 000000000058: 30 32 36 66 37 65 026f7e Haven't heard much issues from the Outlook users (even our OA users) other than the OA users saying they get a random "exchange server not available" message for 15-20 seconds before getting an "exchange server available again" message.

Add me to the list, been working with support for 2 weeks and 0 progress. Noticed I believe Ryan in this thread suggested renaming davex.dll, which I was going to try only to realize it does not exist. I am on Server '08 Exchange 07 SP1 rollup 3.

Renaming the dll works as an interum fix. Sam, you really should look a little harder for that bad boy. I'm running the same setup as you, and I found it at C:\Program Files\Microsoft\Exchange Server\Bin\davex.dll.

Are any of you using blackberries to sync or entourage? It's when those devices hit our server that owa crashes. I've got a case open with microsoft and there are devs working on this issue.

I just started getting this error 2 days ago. Our CAS was having no problems before this, and our OWA was up for over a month solid without any problems. What I did 2 days ago that probably is related to this problem, is upgrade our backup exec remote agents to version 12 and installed the backup exec Microsoft Exchange remote agent. Windows Server 2008 x64, 24 GB of RAM with Exchange 2007 sp1, all updates. Anyone else getting this problem, also running backup exec?

Not using backup exec.

Where is Microsoft at with this thing? I have the same problem. Just happened out of the blue on a Saturday (no changes made). Davex and IIS Worker process crash consistenly every 15 or so minutes.

I'm going insane with this as well. I too have an open case with MS regarding it (opened probably 2 months ago), and its allegedly under development. They won't give me an ETA on it.

BTW I'm not running Backup Exec, but do have a remote agent installed for backup. We use around 50 BlackBerries and no ActiveSync devices. Everything was fine in a test environment with around 5 mailboxes. I transfer all of my remaining mailboxes (around 100) and bam, this starts happening. I'm having trouble even getting OWA restarted at this point. MS needs to get on this.

any luck with this yet. Having the problem using blackberry or entourage. If i use https:\\exchange.domain.com\owa, i connect but don't synch mail if i use https:\\exchange.domain.com\exchange\user@domain.com or just exchange.domain.com then i get the exprox.dll file crashing and it kills the MSExchangeOWAAppPool in iis. same 2k8/2k7 config. entourage worked fine when i was using a 2k3/2k7 config. using a 2 cas/hub server and 2 mailbox in ccr config. updated all the servers to update pack 3 I ve tried a million thing and havnt had a single breakthough yet

Received this email over the weekend. Hopefully it means they're close. Hi Mathew, I wanted to give you an update on this. Currently Exchange SE team is working on porting the fix to Exchange 2007. I will be sending you the fix once they finished testing and is released. Let me know if you have any questions. Regards, Brunda

Is there any update on this? We're experiencing the same issue and waiting on the fix as well. Thank you, Chris

We worked with some devs last week. Turns out a redirect we had programmed in, IIS couldn't handle it, even though the instructions we got came right from microsoft. The redirect basically directed people who would go to the wrong url it would take them to OWA. They had us take that out and then OWA would work. Once we got OWA working though none of our blackberries could sync with Exchange 2007. We don't use BES they use a wireless sync through verizon. Microsoft did confirm it is a bug and they need to release a fix so it can handle the redirect. So if you're doing anything like that in IIS take it out.

Found the following article that seems to address this issue head on.Simply changing from a Redirect to point to a static File that has a html redirect stabilized it for us. Take a look. Hope this helps. http://blogs.technet.com/mblant/archive/2008/08/29/exchange-2007-owa-crash-issue-on-windows-2008-server-iis7.aspx

So let me get this right. in IIS 7 settings You setup a http Redirect at the default website level to https://exchange.domain.com/owa then you set up the file to point to that same address and set it up for a response to 403.4 That works for me as far as redirecting but then entourage doesn't crash the server but it also doesnt authenticate, and i cant check my active sync systems. The thing that throws me is that when you set an http redirect at the root it propogates down and you can't uncheck it on the \exchange folder.

We have recently deployed Exchange 2007 w/SP1 w/Update Rollup 3 on Server 2008 a few months ago and are also having the same issue where the applicationpool 'MSExchangeOWAAppPool' crashes. It seems to be related to the exprox.dll according to our application event log.I have read this entire post and went to the different links and still don't see a real solution. Do you think that by moving the following virtual directories "/exchange", "/exchweb", "/public" into their own application pool would at least solve the issue of OWA crashing? If so, are there any negatives with moving the mention virtual directories to their own app pool? Thank you for your assistance.

I finally broke down and called Micrsoft Product Support team and it turns out that Entourage clients are causing the MSExchangeOWAAppPool to crash. When a message size greater than 48K is sent, the app pool crashes. Microsoft has an interim fix if you are having this problem. I was told that Update Rollup 5 for Exchange 2007 SP1 will have this fix included. If you want the fix, you must contact Microsoft Product Support. This was the title of my case "Exchange 2007 OWA App Pool Crashes (Exprox.dll to blame)". -Robert

Im having the exact problem with our entourage users. Has there been a resolution to this? any help would be greatly appreciated.

I have this same issue and have been troubleshooting it for weeks. I thnk it started on Rollup 3. I get this error reported everytime someone tries to log into OWA. After entering credentials, OWA alerts user session timeout and i get an error report saying "'MSExchangeOWAAppPool' suffered a fatal communication error with the Windows Process Activation Service." Any updates on the patch or fixes available will be much appreciated.

I think the previous fix worked for those of us without entourage - removing the http redirector from the default website and using a file instead. As previously posted this article http://blogs.technet.com/mblant/archive/2008/08/29/exchange-2007-owa-crash-issue-on-windows-2008-server-iis7.aspxdescribes in detail why the http redirector crashes the owaapppool.

Hey matthias1977, I have tried the post methods of the blog you specified and still there is no success for me. Is there anyone else that just has just Server 08 and Exchange 07 SP1 with my same issue rather resolved or not. I really need to know if this is something that the Exchange Team is working on.

So you've eliminated all http redirects in your default site and all virtual sites? You have to get rid of them all, as Exchange doesn't understand them. As far as the Exchange team working on it, I've had an open case with MS for months and haven't been updated in probably a month. A fix was allegedly under development, but like I said I haven' t heard from them recently regarding it. You can call MS and open a case. They won't charge you since it is a flaw with Exchange 2k7 and server 2k8 - well they will but they'll refund the money.

This also worked for me.greg Ryan VI wrote: Renaming davex.dll to .old worked for me. This was the only post that suggested this Thanks...

Has anyone got the hotfix/KB number for this?I can't find any relevant kb articles, so Microsoft Product Support want to charge me before they'll let me speak to anyone technical.For the record, we have exactly the same problem with Server 2008 x64/Exchange 2007 SP1 Rollup 3 and Entourage clients. Best wishes,Rob Robert Lightner wrote: I finally broke down and called Micrsoft Product Support team and it turns out that Entourage clients are causing the MSExchangeOWAAppPool to crash. When a message size greater than 48K is sent, the app pool crashes. Microsoft has an interim fix if you are having this problem. I was told that Update Rollup 5 for Exchange 2007 SP1 will have this fix included. If you want the fix, you must contact Microsoft Product Support. This was the title of my case "Exchange 2007 OWA App Pool Crashes (Exprox.dll to blame)".

I just sent off an email to see if there is a hotfix #. I'll keep you posted.

Evidently an interim update is different than a hotfix. He indicated there is no hotfix, but that I could request an interim update. I know they are not charging me for my incident, since it is a flaw in there system. You'll have to pay upfront, but they should refund your money to open a case with them and request an update. Also, I don't know if he misspoke, but he indicated it would be in rollup 6: There is no hotfixfor Exchange 2007. Note: For better understanding of the Exchange 2007 servicing model, please see the following page. Fix is targeted to be released with Rollup Update 6, currently there is no Interim update developed for this issue. Let me know if you want me to request for an interim update.

Matthias, thanks for the useful info.As a temporary fix we've changed the Maximum Failures (under Application Pools > MSExchangeOWAAppPool > Advanced Settings) to 20 failures in a 5 minute interval (up from the default 5 failures per 5 mins).We only have a handful of Entourage users at the moment, so unless they all connect at once, we shouldn't see the WAS 5002 error, where the "Application pool 'MSExchangeOWAAppPool' is being automatically disabled".Other than that, we'll just have to wait 6 months or so for rollup 6 to arrive - it's back to using OWA for most of our Mac users then!

Robert Lightner wrote: I finally broke down and called Micrsoft Product Support team and it turns out that Entourage clients are causing the MSExchangeOWAAppPool to crash. When a message size greater than 48K is sent, the app pool crashes. Microsoft has an interim fix if you are having this problem. I was told that Update Rollup 5 for Exchange 2007 SP1 will have this fix included. If you want the fix, you must contact Microsoft Product Support. This was the title of my case "Exchange 2007 OWA App Pool Crashes (Exprox.dll to blame)". -Robert Hi all, Please try to identify that your issue is like: When sending a large message from Entourage to Exchange 2007, the w3wp.exe crashed on exprox as always. It should be kb 956319 which has not been released, you may contact PSS to get it. So far ,this hotfixwill be included in rollup 5. If it is not the case, then I recommend you to contact css to get better help. Hope it helps. Xiu

For all your 2k8 server / Exchange 2007 / Entourage client environment supporters out there, here's what worked for me: - Search your disk(s) for anotherexprox.dll. I was able to find Version 8.1.240.5 (November 20, 2007) in one of the $PatchCache$ subdirectories in C:\Windows\Installer. - Replace your current exprox.dll in the \Exchange Server\bin directory with the older version (naturally, don't forget the backup copy first) Replacing a dll with an older version is of course an absolute "don't do" under normal circumstances, but after having lost a lot of sleep over this, it was definitely worth a try. And it worked for me... Entourage no longer crashes the App-Pool... unless (here comes the BUT) you try to access Public folders...

It would be nice if you actually gave a KB article that existed. KB956319 does not exist, please provide a real article. Thanks. Xiu Zhang - MSFT wrote: Robert Lightner wrote: I finally broke down and called Micrsoft Product Support team and it turns out that Entourage clients are causing the MSExchangeOWAAppPool to crash. When a message size greater than 48K is sent, the app pool crashes. Microsoft has an interim fix if you are having this problem. I was told that Update Rollup 5 for Exchange 2007 SP1 will have this fix included. If you want the fix, you must contact Microsoft Product Support. This was the title of my case "Exchange 2007 OWA App Pool Crashes (Exprox.dll to blame)". -Robert Hi all, Please try to identify that your issue is like: When sending a large message from Entourage to Exchange 2007, the w3wp.exe crashed on exprox as always. It should be kb 956319 which has not been released, you may contact PSS to get it. So far ,this hotfixwill be included in rollup 5. If it is not the case, then I recommend you to contact css to get better help. Hope it helps. Xiu

The fix is included in Update rollup 5. http://support.microsoft.com/kb/kb956319 Cheers.

Yes, Rollup5 for Exchange Server 2007 sp1 has been released. http://support.microsoft.com/kb/953467/ You can download it from http://www.microsoft.com/downloads/details.aspx?FamilyId=652ED33A-11A1-459C-8FFE-90B9CBFE7903&displaylang=en Regards, Xiu

I don't know about you all, but RPC over HTTP/Outlook Anywhere(also running through IIS7's app pools) seams to be getting more and more unstable with each update rollup. I've got two different boxes in two disjoint domains with the same issue...having to reset IIS periodicly for the clients to quit getting prompted for authentication repeatedly. I suspect it's the same issue. I used to be able to make the problem go away with the davex.dll hack, but with update rollup 5, nothing helps anymore.

Hi, I notice there seems to be a couple of issues logged here. both to do with MSExchangeOWAAppPool crashing. 1) Windows 2008 (iis7) - redirect causing crashes 2) Exchange 2007 SP1 - Entourage and messages above 48K. Well I dont seem to fit into any of these categories.Here's my issue: All Servers Windows 2003 SP2 All Exchange Servers 2007 SP1 We have two CAS servers and 3 Exchange MailboxClusters. Overthe last weekboth of ourCAS servers have come up with a Serivce Unavailable error on OWA.Once each and at different times. Having investigated the Event Logs I see the following: Event Type:WarningEvent Source:W3SVCEvent Category:NoneEvent ID:1011Date:28/01/2009Time:23:32:40User:N/AComputer:EXCASUM01Description:A process serving application pool 'MSExchangeOWAAppPool' suffered a fatal communication error with the World Wide Web Publishing Service. The process id was '5888'. The data field contains the error number. For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp.Data:8007006d and after 5 in 5 minutes we get: Event Type:ErrorEvent Source:W3SVCEvent Category:NoneEvent ID:1002Date:28/01/2009Time:22:52:33User:N/AComputer:EXCASUM01Description:Application pool 'MSExchangeOWAAppPool' is being automatically disabled due to a series of failures in the process(es) serving that application pool. For more information, see Help and Support Center at http://go.microsoft.com/fwlink/events.asp. The 1011 errors have only started in the past week and don't seem to follow any pattern. I looked atthe article:http://support.microsoft.com/kb/885654A user receives a "Service Unavailable" error message when the user browses your Web site that is hosted on IIS 6.0 This is a general article and the OWAAppPool doesnt run under the Network Service Account so it doesnt seem to be the issue. We aren't using windows 2008 so the IIS redirect cant be the issue. We do have Entourage users but we dont have any of the following events in the application log: Log Name: Application Source: Application Error Event ID: 1000 Task Category: (100) Level: Error Keywords: Classic Description: Faulting application w3wp.exe, version 7.0.6001.18000, time stamp 0x47919ed8, faulting module exprox.dll, version 8.1.240.5, time stamp 0x47427a9a, exception code 0xc0000005, fault offset 0x0000000000068511, process id 0x%9, application start time 0x%10. Because of this I would suggest our crash isn't being caused but the 48k message issue outline inhttp://support.microsoft.com/kb/kb956319. The two outages have happened out of hours (between 11pm and 4am) Does anyone have any idea's as to what is causing my OWAAppPool crash. Regards, Mike Mike Loveless