View Single Post
Old 08-01-2006, 08:18 AM   #9 (permalink)
Sagz
Knows Where the Search Button Is
 
Join Date: Feb 2006
Model: 8100t
Carrier: tmobile
Posts: 45
Post Thanks: 0
Thanked 0 Times in 0 Posts
Default

I like to use v-file to read my logs cause I can do a search across all logs. www dot fileviewer dot com. It also splits the logs up so you dont eat your memory up while viewing.

Some basic maint things you can search the magt log are
"Optimize Threadpools" = This one will tell you the amount of threads allocated to each connected mail server along with how many handhelds connected. This is very handy for load balancing and preventing hung threads.
Here is an example of what you would see:
Optimize ThreadPools, total number of users 1117
[30000] (08/01 00:13:51):{0x250} No empty thread-pools were found.
[30000] (08/01 00:13:51):{0x250} Before optimization: total number of worker threads 128, where 54 non-removable
[30000] (08/01 00:13:51):{0x250} Thread pool for mail server (blahblah/srv/blah) has 3 threads to serve 30 handhelds
(blahbla/srv/blah) has 32 threads to serve 81 handhelds


Next handy thing you can search for in MAGT is: "GetConnection"
Sample return:
[30000] (07/99 13:48:48):{0x12F4} ConnectionItem::GetConnection(): Connection wait time exceeded 100ms threshold. Time spent waiting was 30125ms. There are 91 thread(s) still waiting

Or do a "ODBC" Search if things are bad you may see something like:
[40000] (07/32 01:03:22):{0x13A0} [ODBCConnection::Connect] DB=SQL Server, LoginConnectionTimeOut=30

This reflects hung threads. To troubleshoot, see if you can connect to your SQL db. Look to see how much pending there is and which mail server connections may be related. See if a mail and cluster server are both down at the same time.


Most of the time for MAGT logs Im doing user searches.


In the alert log depending on the level of your logging you can do house cleaning by pulling the invalid address books, journals and so forth that get logged in there.

In the dispatcher log look for things like:
"COM Error" if you are having think you have or want to pinpoint a connection problem it will return things like:
[10226] (07/34 14:01:29):{0x1EC0} COM Error 0x45BFF0 in CheckDB - [DBNETLIB][ConnectionWrite (WrapperWrite()).]General network error. Check your network documentation. - Unspecified error
[10224] (07/32 14:01:29):{0x1EC0} Database error in CheckDB (err=0x80004005, native err=11) - [DBNETLIB][ConnectionWrite (WrapperWrite()).]General network error. Check your network documentation.

or
[10213] (07/00 14:03:10):{0x1EC0} OpenDB: COM Error (80004005): Unspecified error.
[10251] (07/23 14:03:10):{0x1EC0} Failed to authenticate connection - local machine time may be out of sync with domain controller time. Please restart Windows Time Service.
[10214] (07/54 14:03:10):{0x1EC0} OpenDB: OPEN_DB failed.
[10218] (07/65 14:03:10):{0x1EC0} CheckDB: Could not open database.



Hopefully that helps I could go into more detail if needed, but this is a long post.

Last edited by Sagz : 08-01-2006 at 08:21 AM.
Offline