SP1 Install Caused Problems
Had a couple questions about some Event ID 6702 errors on the server I’ve been getting recently.I can’t seem to find any intelligible information on Microsoft’s site about it; however, I did find this:
http://support.microsoft.com/kb/842006/en-us#XSLTH5294121122120121120120
There are no replication partners for all I know on this server.As far as the A records that it wants me to check, when I take a look at dnsmgmt -> Forward Lookup Zones -> DI-COMM.local, there are several A records for all the machines on the network.The support doc is telling me to delete all A records except those corresponding to the actual server.As far as I know, I NEED these A records for the name lookup to resolve to IP.What is this doc trying to tell me?Again this was another MS generated problem after the install of SP1.The site actually tells me to just ignore the error if there are no replication partners for the server.The server has 2 IP addys, only one used for any type of serving (default setup to use 192.168.1.200 for actual serving).Should I be getting rid of the 192.168.1.201 record?
Secondly, “sqlservr” service is using way too much memory lately (1.2GB peak so far), considering it is not actually serving any databases.The wsus is running its own msdse mini sql server, and I can’t think of anything else that would actually be using the MS SQL Server.I went to the services msc to see if I could stop/restart the service to see if it would clear anything, however the only services I see even relating to SQL are:
[Status is started, startup type is automatic]
MSSQL$SBSMONITORING
MSSQL$SHAREPOINT
MSSQL$WSUS
[Status is blank, startup type is manual]
MSSQLServerADHelper
…
[Status is blank, startup type is manual]
SQLAgent$SBSMONITORING
SQLAgent$SHAREPOINT
SQLAgent$WSUS
So I don’t actually see a “sqlsrvr” service.
Lastly, I have setup monitoring and reporting (obviously since I’ve sent you forwarded e-mails), but I will get like 3 copies of the same exact email (for instance it will send me 3 e-mails for a performance report for 2/27/06 6:00 AM – all the exact same e-mail).Again I believe this started after the sp1 install.Man do they ever have bugs.There support site has a lot of sp1 related issues that I’ve seen, I’m just not able to search for exactly what I need….
If you’ve got some knowledge on these subjects, or have the time to point me in the right direction I would greatly appreciate it.Thank you,
Mark Cass
Systems Technician
Diversified Communications Inc.,
D.B.A. DI-COMM
E-mail:mcass@di-comm.com
[6606 byte] By [
mcass] at [2008-2-21]
Yes and NO. I found out that the memory hog is the monitoring and reporting service. If left unattended (as in if I don't stop/re-start the service every couple of days, or restart the server), the sqlservr.exe used for the MS monitoring and reporting will eat and eat and eat and eat RAM. I just took a look, and that process is eating 1,010,768KB at the moment. 1GB of RAM for monitoring and reporting?!?!?!?! This has to be some kind of memory leak or something. If it's not, then what are you supposed to do?...restart the server every day?... and our server is not a heavily used one at that! I wonder what a larger company's heavily loaded server will do?
This has to be related to an SQL issue in some sort of way, but I don't know how.
ok.....after going to a MS TechNet Event in Fayetville, AR. A nice gentleman offered me a get-out-of-jail free card. I called Microsoft support using this card and they fixed my problem. The multiple copies of e-mails was my fault. Apparenlty since I had 3 different rules in my Outlook for sending the performance reports to one folder, the usage reports to another, and alerts to yet another, the multiple rules gave me the multiple copies of the same e-mail....easy fix...just one general rule for everything coming from that e-mail addy or with your server name in the subject line. The major problem (BIG MEMORY LEAK) could only be fixed in a round a bouts way.
AND IT IS AN SQL PROBLEM. Maybe SQL is not necessarilly the cause of the problem, but the integration between the monitoring and reporting and the sql server is.
The tech support lady had to remote into my server and change several sql variables to limit the size of the database for the monitoring and reporting to only 200MB. I say this is only a round a bout way of fixing the memory leak symptom of the problem, because limiting the size of the DB does not fix the cause, only the symptom. hmmmmm.....kinda sounds like doctors doesn't? So, now instead of 1.2GB, my DB maxes @ 200MB, and it will stay there all the time.
Sorry that I cannot provide you with the exact details of how to perform this, as it was done fairly quickly by remote (as if this is a known issue and they already have a temp fix for it). However, if you happen to know someone fluid in sql, or that at least knows some general basics, they might be able to help you out further on performing a DB max size limit.
I have the same problem too.
MSSQL$WSUS and MSSQL$SBSMONITORING eat up all the memomy until there is none left. I have 4GB of memory and those two services made the server use the swap file because it was out of memory. One was using 1.6GB and the other 1.2GB.
I think Microsoft should come out with a tool or an official patch to limit the amount of memory those two process/services use.
One of my friend told me he could limit it with the SQL Server 2000 Admin tools. We'll do it this week.
But I still think Microsoft should address the issue.
