HELP! with server 2003 2

[joeblk]

Hi All.
Could somebody please help me with this. I have a Dell machine running 2003 server that is grinding to a halt. The machine is almost non-responsive. Lately it takes forever to do anything on it.

The machine has been fine up until last week with no apparent reason. I have not installed anything on it, not even updates.

It started with one of my remote users complaining that they are sometimes not able to log on to the machine (this symptom is intermittent) and then I checked and also was unable to log on to it. It occasionally came up with a messge: the connection to the remote computer was broken. This may have been caused by a network error. Please try connecting to the remote computer again.

The affected machine is:
Dell PowerEdge 1800
OS: 2003 Server
RAID 5 configuration

My setup is as follows:
User VPN's unto the network and then use remote desktop to log on to this machine (call it computer B). No trouble VPNing to the network, both from external and internal. I also have four other machines on the same subnet and they are not experiencing this problem. RDP takes forever to open up and opens the window in "blocks" which could take up to 3-5 minutes. At other times it comes up with the "the connection to the remote computer was broken" message and refuses to connect. When it manages to connect, it takes unacceptably long to open applications, even bringing up properties page takes > 10 seconds.

I am running the same application/s on the other (identical) machines on the same subnet and they are OK.

What I have tried so far:
1. I have changed the background selection in RDP to noe and this seems to have improved the connection slightly (NO MORE the connection to the remote computer was broken message), however the problem remains when accessing data.
2. changed MTU to 1360 in the registry
3. reset TCP/IP stack using netsh.exe utility (this did not produce any log - not surprising since nothing has changed on the machine)
4. rebooted like a million times
5. Pinging the machine gets a reply

The performance has not improved and I have practically ran out of ideas and need your assistance/suggestions to try and resolve this.

Please HELP!

Thanks
Joe

Could this be Hardware issues?? even though nothing in the event log suggests any problem?

 

[porkchopexpress]

When this is happening is the server unresponsive if you log on locally?

 

[joeblk]

Yes, logging on locally is fine, but still takes as long to access data.

New development since I last posted.
I have tried to ghost the machine (just in case it all goes peer shapped), but ghosting a single partition of 12 gb says it will take >81 Hours (eighty one hours). It has done 1 and half hours and still says it has 80 hours remaining.

Pleaes help and let me know if you require any more information.

THANKS a million.

Joe

 

[rjs]

Just a thought, but check the RAID status. Perhaps one or more of the drives have failed and you are running in a degraded state.

Also look at the event viewer for any errors.

You could also check task manager. Make sure you check the box to show processes for all users.



R.Sobelman

http://www.tracklogins.com

 

[joeblk]

Porkchopexpress, Thanks for your contribution, please find my response above.

RJS, I did check the event log and it had nothing that looked relatively connected to this issue. I will have to go into work tomorrow since I have left the ghost running overnight. I can post the event log messages if that would help.

Also i did check at boot up and no raid error came up. Correct me if I'm wrong but I expect to see some disk or raid message if there is a disk failure. (is there anywhere else I should be looking)?

Thanks and we keep trying.

Joe

 

[markdmac]

Try this (note requires restart):

'==========================================================================
'
' NAME: MSKB914052.vbs
'
' AUTHOR: Mark D. MacLachlan , The Spider's Parlor
' URL: [URL unfurl="true"]http://www.thespidersparlor.com[/URL]
' DATE  : 4/20/2006
'
' COMMENT: Repairs registry entries that when missing cause Windows 2003 servers to behave very sluggishly.
'
'==========================================================================
On Error Resume Next
Dim path, WSHShell
Set WSHShell = Wscript.CreateObject("WScript.Shell")
path = "HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\Winlogon\Notify\termsrv\"
WSHShell.RegWrite path & "Logoff","TSEventLogoff","REG_SZ"
WSHShell.RegWrite path & "Logon","TSEventLogon","REG_SZ"

If Err Then
	MsgBox "Error Encountered"
	WScript.Quit
Else
	Report = "Setup Successful"
End If

mname="."

If Msgbox(Report & vbCrLf & "You must reboot to complete setup." & vbCrlf & "Do you want to reboot machine now?", vbYesNo, "Reboot Machine") = vbYes then
		Set OpSysSet = GetObject("winmgmts:{impersonationLevel=impersonate,(RemoteShutdown)}//" & mname).ExecQuery("select * from Win32_OperatingSystem where Primary=true")
		for each OpSys in OpSysSet
			OpSys.Reboot()
 		next
end if

I hope you find this post helpful.

Regards,

Mark

 

[joeblk]

Hi All,

Thanks for your suggestions so far.

I have looked in the event log and thre are no warnings or errors relating to the problem. Actually the only errors there were about incorrect log ons and a few about printing.

I have also opened task mgr and tried to open as many applications as possible but the CPU usage mainly stays at 0% - 11%. (now please correct me if I'm wrong but I would expect the usage to be a lot higher than that).

MARKDMAC
How do I run the above script? Should I just copy and save it to any location to run?

Thanks all for your contribution so far.......

Joe

This makes me think that maybe I have a "memory leak"? what do you guys think please? and if that is the case, how do I rectify it?

 

[markdmac]

Save the above script to a text file. Name it MSKB914052.vbs or anything with a .VBS extension. Then double click it to execute.

Memory leak is certainly possible, however I would rule out the issue I have pointed out first as I have seen this behavior.

I hope you find this post helpful.

Regards,

Mark

 

[joeblk]

Hi Markdmac,

I have ran the script but it has not made any difference to the performance. I also increased the paging file but still no joy.

Any more suggestions on what I should check or try please.

Thanks
Joe

 

[markdmac]

I'd run through the list of usual culprits to start.

1. Update your AV and do a full scan.
2. Run anti-spyware software and make sure there is no spyware.
3. Check running processes in Task manager. See what processes are hogging either memory or CPU.

I hope you find this post helpful.

Regards,

Mark

 

[joeblk]

Thanks, I did update and run AV on it on Saturday but nothing found.

I also checked task manager and there is no obvious process hugging up memory. In most cases the CPU usage stayed at 11% at the most.

I even increased the minimum page file size but stil no joy.

Kind Regards

Joe

 

[markdmac]

Take a look at PMON and see if it can give you any more indication of what is going on.

http://www.sysinternals.com/Utilities/PMon.html

I hope you find this post helpful.

Regards,

Mark

 

[mrslimey]

Someother things you may wish to check.

How much space is left on your system partition especially now that you have increased the page file?

Do you have the Dell OpenManage Applications - Array Utility installed.
Usually this is found in Start-Programs-Dell OpenManage Applications-Array Manager-Array Manager Console.

From within this utility you can clearly see whether your disk is degraded.

Other checks - how much memory do you have installed and number of processors?

Have you done the usual boring checks of defragmentation ? -not likely to have a great impact but never the less always a good check.

In the Windows event logs you will be wanting to keep an eye on the system events. Application and Security won't be of much help in this situation.

Have you added any other servers or applications to the network as a whole since you started seeing problems with this server.

 

[joeblk]

Hi All,

Sorry i have been away on a two day conference and only just came back.

Just a quick update from the last time.
After trying a few more troubleshooting AGAIN, I finally got a RAID message that it is running in DEGRADED mode. Now I don't know if this is consequential of what I have done yesterday or the original problem since this has been going on for over a week now (all the while saying it was running in optimum mode)

Message on the screen was: ARRAY # 0 = RAID 5 - DEGRADED.

I went into the Dell utility as advised by mrslimey and it is trying to rebuild the array at the moment. I am not sure if it would rebuild on the same faulty disk otherwise I would need to replace it.


Once again, thanks all for all your suggestions so far.

Kind Regards

Joe

 

[joeblk]

Hi All,

OK, sorry but since the my last message I

1. Configured a temporary machine for terminal services on the same domain as the faulty machine and it works just fine as expected. So I have my users logging on to this for now.

Faulty machine:
2. Got an error on the RAID Array Manager Console that the "disk #2 is offline" I got in touch with the hardware supplier and they sent in a replacement disk which I used to rebuild the array yesterday. No RAID error messages has appeared so far.

The machine is still as slow and RDP still will not connect, so the above has not made any difference.

I have checked the windows system event and it has the following error codes:

Code: 4321
The name "....." cld not be registered on the interface with IP Address "1.1.1.1" (Local machine's IP address). The machine with IP: 2.2.2.2 (the domain controller's IP) did not allow the name to be cliamed by the machine. This message has been appearing way longer that the problem. The first of such errors was in Novemver 05.???

Also

Error code: 111666
The system faile dto register host (A) resource record (RRs) for network adapter with settings: ..........

I was holding out hopes that changing and rebuilding the RAID will resolve this but it has not.

Any more pointer will be greatly appretiated.

Thanks All

Joe

 

[joeblk]

Hi Again,

Just a quick update. As per my previous messge, I have rebuilt the Raid and there was no noticeable difference in perfomance. However a week later the machine seems to have come back to live. The performance has increased and it is now working as expected.

I had not done anything else after the rebuid so I guess it took that long to "catch" up.

Thanks again Markdmac and mrslimey

 

[dberg35]

Just a little after thought, is the machine still under warranty? If not you might want to consider replacing it because whos to say it won't happen again. I run a Dell PowerEdge 2800 and it has been rock solid.

 

[joeblk]

Hi dberg35, cheers I'm actually talking to Dell at the moment about possible replacement.

 

[dberg35]

What model are you looking at?