Strange DCOM errors every day

[irbk]

Every day at the same time on 4 out of my 11 DC's (it's the same time on each day, but not the same time on each server, for example server1 might get the error at about 9:30am every day but server2 gets it at about 2:30pm) I get the following errors.

This error showing 14 different clients:

DCOM was unable to communicate with the computer <%computername%> using any of the configured protocols.

and this error showing 18 different clients (these clients are also different from the 14 different clients above)

DCOM got error "2147944122" from the computer <%computername%> when attempting to activate the server:
{8BC3F05E-D86B-11D0-A075-00C04FB68820}

While some servers show clients that other servers don't, for the most part it's the same 32 different client machines that are mentioned in the errors every day. (For example, client machine JSmith may be mentioned as an "unable to communicate" in server1 @ 9:17 am PST, server2 @ 2:11 pm EST, and server4 @ 12:20 pm CST but not on server3 etc) Note that if a client machine is mentioned as "unable to communicate" it won't be mentioned as a client with a "2147944122" error, even on a different server. So using the above example, if JSmith is "unable to communicate" in server1 it will show as "unable to communicate" in server2 and server4 as well. I do NOT ever see <%computername%> "unable to communicate" in server1 and then <%computername%> as a error "2147944122" on a different server.

I've used Mr. Google trying to get rid of this error for about 2 weeks now. I can't figure out what the problem is. The remote administration firewall rule is on, TCP port 135 is open on incoming connections. Heck, I can't even figure out what is different about the 4 DC's that show the error compared to the other 7 DC's that don't show the errors. Most of the machines mentioned in the above errors are systems that only connect to the network through VPN. However, this is not the case for all of the clients.

The only thing that sort of seems to make since is it's remote servers that throw the error. Using the example above, the JSmith error thrown by server1 is in CA, server2 is in TN, and server4 is in WI however the error for JSmith doesn't show in server3 which is at JSmith's location in IL but server3 does show errors for others.

I'm totally stumped. Any suggestions?

 

[technome]

This no fix but I get COM errors from a few sources on some servers. I fixed them multiple times, but but they come back, anywhere from a week later to month. Now I live with them, as long as there are no functions affected.



........................................
Chernobyl disaster..a must see pictorial

http://www.kiddofspeed.com/default.htm

 

[irbk]

Ok, now I'm starting to see the same errors on NON-DC systems too. Any suggestions please!?

 

[irbk]

I did a little experiment today. On my server in CA I'd get the errors every morning at ~ 9:30am PST.
While I wasn't sure about the timing, I felt that the DCOM errors started about the same time as an update for McAfee VS 8.8 was applied. This morning I uninstalled VS from the CA server. 9:30am PST rolls around and guess what...
NO MORE DCOM ERRORS!

This afternoon I installed VS 8.8 again. I'll see tomorrow of the DCOM errors come back. So far it kind of looks like the DCOM errors were being caused by McAfee.

 

[irbk]

DCOM errors came back after I re-installed McAfee. I think that's pretty conclusive evidence that McAfee VirusScan 8.8 is causing the errors.

 

[technome]

Like I stated, live with the errors. You could spent weeks attempting to fix them before giving up, and none of the software producers will help repair the errors.


........................................
Chernobyl disaster..a must see pictorial

http://www.kiddofspeed.com/default.htm

 

[irbk]

Oh, when I tell them "fix the problem or I'll stop using (and paying for) your software" suddenly they get a LOT more helpful. This isn't my first DCOM error dance with McAfee.