eseutil and operations of
I have 4 message stores on my system with each being about 30 gigs in size. I take one message store offline at at a time to do compaction on it which minimizes having my whole organization offline. We run 7 days a week 24 hours a day.
Problem I am having is that the eseutil is really processor intensive and causes other mailboxes to occasionally not respond and other problems related to eseutil hogging all the time or hard drive or something. When I look at the task manager performance chart I see all 4 of my cpus and it usually shows the last one as having more processing going on while eseutil is running while the others are thier normal minimal usage. So I don't think it is processor problem but not sure of anything.
What can I do to be able to run eseutil and still have my other applications/stores run ?
thanks
Problem I am having is that the eseutil is really processor intensive and causes other mailboxes to occasionally not respond and other problems related to eseutil hogging all the time or hard drive or something. When I look at the task manager performance chart I see all 4 of my cpus and it usually shows the last one as having more processing going on while eseutil is running while the others are thier normal minimal usage. So I don't think it is processor problem but not sure of anything.
What can I do to be able to run eseutil and still have my other applications/stores run ?
thanks
ASKER
exchange 2003. I don't run it regularlly. As a matter of fact this is the first time in 10 months I have run it. I first realized that the online does not do much when I had a 160 gig store and took 98 percent of users out of it but the store after several days was still around 160 gigs.
So I took that one offline and did offline compaction which then turned it into a 12 gig store. So I think that I should either quarterly or semi anually according to how many deletions and additions of mailboxes I have do one.
I think running it from another machine will take what I have found is a 4 hour process on my 30 gig stores really really take a long time as they are just 100 meg etherlinks.
So I took that one offline and did offline compaction which then turned it into a 12 gig store. So I think that I should either quarterly or semi anually according to how many deletions and additions of mailboxes I have do one.
I think running it from another machine will take what I have found is a 4 hour process on my 30 gig stores really really take a long time as they are just 100 meg etherlinks.
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
I thought of those items but of course not my preference. I was hoping thier was some kind of "lower" priority that can be set to our command line eseutil to free up resources for other uses.
Not with ESEUTIL,
You would need a 3rd party solution, like this; http://www.lucid8.com/product/preventative.asp
That isnt cheap though... However, from the limited information I have about your environment, it may be beneficial...
-red
You would need a 3rd party solution, like this; http://www.lucid8.com/product/preventative.asp
That isnt cheap though... However, from the limited information I have about your environment, it may be beneficial...
-red
SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
above is definitly an option if I don't find a better one. I guess that was one of my questions if I have errors in database as well as compaction. will these get drug over to new store also?
Greetings:
Do you have more than a gig of memory of the machine(s)? Here's a M$ article for optimizing.
http://support.microsoft.com/kb/815372/
If you do move the mailboxes, be aware of some of the limitations with the recovery storage group. Any serious changes involving active directory will make recovery storage impossible to use on a production system.
http://www.microsoft.com/technet/prodtechnol/exchange/2003/library/ue2k3rsg.mspx
regards
Do you have more than a gig of memory of the machine(s)? Here's a M$ article for optimizing.
http://support.microsoft.com/kb/815372/
If you do move the mailboxes, be aware of some of the limitations with the recovery storage group. Any serious changes involving active directory will make recovery storage impossible to use on a production system.
http://www.microsoft.com/technet/prodtechnol/exchange/2003/library/ue2k3rsg.mspx
regards
While the eseutil is running you could right click the process in task manager and set the priority lower if that is what you want to achieve.
Still not sure what the actual question is.
Still not sure what the actual question is.
ASKER
simonpainter
that sort of is my question. can i set it lower were it will actually use less cpu/disk time so my other applications will continue to run.
that sort of is my question. can i set it lower were it will actually use less cpu/disk time so my other applications will continue to run.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
That will only affect cpu usage though, right Simon?
I think this will still give you the problem detailed earlier (by doing it over the network) - it will take a lot longer (relative to what you set it to)
You would think that with 4 processors, it shouldn't matter too much - what else is this server running??
-red
I think this will still give you the problem detailed earlier (by doing it over the network) - it will take a lot longer (relative to what you set it to)
You would think that with 4 processors, it shouldn't matter too much - what else is this server running??
-red
Moving users to another store is a good idea as they can be done one by one at sheduled times so only one user is offline at a time (and it can be scheduled to start out of hours).
Why is there defragging going on over the network? That's just plain dumb!
Why is there defragging going on over the network? That's just plain dumb!
Greetings:
Our servers are so full, we had no choice but to defrag over the network. But with 16 plus information stores ... six servers ... and a few terabytes of data, only a small group of people are offline at a time. -- only about 300 or so.
Our servers are so full, we had no choice but to defrag over the network. But with 16 plus information stores ... six servers ... and a few terabytes of data, only a small group of people are offline at a time. -- only about 300 or so.
Is that six exchange servers?
If so how about moving users to another server while you clear down the stores on the server which is overloaded?
If so how about moving users to another server while you clear down the stores on the server which is overloaded?
So far, the best idea to come out of this thread is this -> http:#17508887
And the validity of defragging over the network isnt what we are talking about here, and there are applications for it, irrespective of the stellar arguments against it so far.
-red
And the validity of defragging over the network isnt what we are talking about here, and there are applications for it, irrespective of the stellar arguments against it so far.
-red
Greetings:
I can defrag an information in less than a few hours. Moving the mailboxes means MORE work and I've just increased my work schedule to days instead of hours. Plus there are a few dozen user calls when they SAID they logged out but didn't. And let's not even discuss how many meetings it takes to convince management the work is necessary and how people will be effected.
It's a personal site decision, isn't it? And my management was happy with a four hour outage to reclaim space. Most of that space was from implementing a new server and moving mailboxes anyway. We run 7/24 as well, and restores immediately after moving mailboxes is painful. Once the active directory structure doesn't match up, the recovery storage group doesn't work. So just randomly moving mailboxes around is not always a good idea. Our users are always losing things and not noticing it for months, so we need to keep the structure as stable as possible.
So .. back on subject ...
I can defrag an information in less than a few hours. Moving the mailboxes means MORE work and I've just increased my work schedule to days instead of hours. Plus there are a few dozen user calls when they SAID they logged out but didn't. And let's not even discuss how many meetings it takes to convince management the work is necessary and how people will be effected.
It's a personal site decision, isn't it? And my management was happy with a four hour outage to reclaim space. Most of that space was from implementing a new server and moving mailboxes anyway. We run 7/24 as well, and restores immediately after moving mailboxes is painful. Once the active directory structure doesn't match up, the recovery storage group doesn't work. So just randomly moving mailboxes around is not always a good idea. Our users are always losing things and not noticing it for months, so we need to keep the structure as stable as possible.
So .. back on subject ...
ASKER
Well lots of ideas here and I will attempt to close and award points properly. My original question was on the system being so busy while moving and how to prevent that. what I got was ideas on various options. The best one would be using another machine to do move and I tried that and it does keep the exchange server from being so busy that it does not process other exchange functions. Of course the real trick here is to get a dos based utility (eseutil) and not have it hog all the system resources.
Thanks for all help
Thanks for all help
You shouldnt be running eseutil on a regular basis - that isnt what it is for.
Exchange will run it's own online defrag each night, this should be enough for any exchange environment.
So, why are you running ESEUTIL regularly (and how regularly) and what version of exchange are we talking about?
If you really do have a good reason for this (and I would love to know what it is) then you could run ESEUTIL on another machine through the network. But, this will obviously increase network traffic.
Hope that helps,
-red