Jump to content

Recommended Posts

Posted

Tarquel

I hope this explains things a little more:-

 

We already have a plan in place to move to Exchange 2010 in the summer. However, between now then we have an issue where the Information Store or the MTA stack stops responding. This is a known issue to Microsoft and they have recommended moving from 2003 !!! The problem is our Head is not prepared to wait in the mornings for me to get in and manually restart the services. So what I need to know is to how to monitor the application event log for errors and restart the offending MS Exchange service automatically.

Posted

Hi

 

Firstly, this know issue you have, was this offcially logged with MSFT?

I have many customers who are using Exchange 2003 and we are migrating them to Exchange 2007/2010 and have not come across this issue.

We can create a several types of dumps for Exchange and try to identify root cause or if you want you can create a script to check the service and start them if they are stopped.

 

Regards

Sukh

Posted
However, between now then we have an issue where the Information Store or the MTA stack stops responding.

 

If those services stop, do u have them set to auto recover i.e. start the service again in the configuration of each service. Being ur on Exch 2003, I'll assume Server 2003 too, and you just need to:

 

Run > Services.msc > Double Click the service(s) in question > Click the Recovery tab and set the Frist, Second, and Third Failure to Restart the Service

 

and see how that goes :) although I'm assuming you've already tried this, or that the service itself doesnt actually stop, and just locks up, meaning that this likely wont work and I could see this being a right pain in the proverbials :)

 

Nath

Posted

Hi Guys

To answer you both in turn:-

1. Those recovery options are already set and yes the service just locks. However, It does generate an event ID, so I was looking for a piece of VB script that will look for the ID and restart my two services.

 

2. As the problem can occur at random, having the batch file will not catch it unless it locks just before the timed restart.

Posted

EDIT: Ignore my post - you've all beat me to my posting hehe

 

@ECHO OFF

CLS

echo - Stopping Services

net stop MSExchangeIS

net stop MSExchangeMTA

 

echo - Starting Services

net start MSExchangeMTA

net start MSExchangeIS

 

Save as:

thestopstartspecial.bat

 

and use Scheduled Tasks (with an account that has permissions to stop those services etc etc)

 

NOTE: The order of stopping them should be checked out first, use ping waits inbetween if needed, etc. I recall Exch 2003 being a bit of a winger regarding that many moons ago. I take no responsibility if it destroys all your data and doesnt take you out on a date to make up for it etc....

 

:)

Posted

With a service that did that sort of lock up, we used procdump [ ProcDump ] with the Microsoft Support Services Engineer, to get the dump files.

 

Cant remember the exact command line to use but it wasnt too hard to figure it out.... might be of use, especially if u want to make a dump of the crash to see whats occurin' :)

 

Nath

Posted

That leads me on to another point Sukh....

 

Is Exchange all updated with SP's and Update Rollups / Hotfixes etc? Can cure alot of wrongs if it isnt... but research any Exch SP's to avoid "surprises" ;)

 

Nath

Posted

Hi

 

@Nathan - yes you can use that tool. There are some other tools specifically to take dumps of store.exe and exchange. Also, depends if it's a crash or in a hung state.

 

I use these tools frequently, however even when I work closely with MSFT some of the dumps have to go to MSFT for analysis as you need the internal code to debug.

 

Cheers

S

Posted (edited)

The actual event ID's you are getting may help us to give you assistance??

 

The scheduled restart of the services may also alleviate the pain you are experiencing from the lockup's as usually with these things the lockups are after it's been running for a while ... so restart services at say 6am and 6pm may stop you running long enough to get the lockup??

 

Have you looked at the Solarwinds Free exchange service monitor, that may help you with getting it up and running or at least getting notified of the failure.

 

Skr

Edited by tarquel
typo :)
Posted
@Nathan - yes you can use that tool. There are some other tools specifically to take dumps of store.exe and exchange. Also, depends if it's a crash or in a hung state.

 

I use these tools frequently, however even when I work closely with MSFT some of the dumps have to go to MSFT for analysis as you need the internal code to debug.

 

Indeed.... I've been there and got that T-Shirt ;)

 

...to help them with this issue: The Exchange Mailbox Replication service crashes when you use the "New-MailboxImportRequest" command to import a PST file to an Exchange Server 2010 mailbox (i.e. importing PST's was a royal pain in 2010 SP1 over RTM....specific service crashing a plenty so gotta say that we, and probably a few other org's, helped get this much improved in U2....tho the recent U3 appears to mention it again, so maybe they've further improved it again :D ....but we digress ;))

Posted

Hi Guys

Here's my plan of attack so far;- I have identified that the error code I want to trap is 9175 and after a couple of lengthy calls to other IT managers. We have stumbled on the following plan which should get us to the summer break and beyond if my MS Exchange budget gets pulled !!

 

1. Download the Eventsmanager tool from GFI

2. Install and point it at the servers application log

3. When the application log has an error and event ID of 9175, restart both MTA Stacks and Information store.

 

It's not ideal, but it will keep us going until the upgrade.

 

Will let you all know what happens tomorrow once it's all set up and running.

 

 

@SkreeM1980: Had a look at the solar winds tool and it didn't give me the information I wanted out of it.

Posted

Hi

 

When do you plan to migrate? How aften does this crash occur? Are able to reproduce or know the timings? Is there a pattern?

 

Do you want to try and identofy root cause?

 

Thanks

Sukh

Posted

Hi

 

The reason why I asked which OS you are using was to see if we can deploy poweshell on your server and write a script to monitor the event id and set a trigger, ie. restart the services.

 

Cheers

Sukh

Posted
@Sukh: The upgrade will happen at the end of the summer term. The OS is 2003 Enterprise R2, There is no pattern to when the services lock, but I do know now that the event id is 9175. Powershell can be deployed onto the server if required.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...