I just blew $1K !!

Woke up this morning and had no idea that I'd be out $1K.

Course, the good news is that it's because Rod Trent emailed me to tell me that I had won a free 5-day pass to TechMentor!

My 1st thought was that I'd have to blow it off because Johan is coming Monday.  But JD on my team told me I'd be nuts to pass up the opportunity, so I'll just miss day one and go anyway.

WOW.  My head is still spinning.

  • Created on .

Upgrade from Server 2008 STD to 2008 ENT

We bought a new server for a child site recently. Our old standard was 32GB boxes. Now it's 48GB. But our standard for OS was STD, not ENT and STD can't see all that 48GB of RAM. And of course, we didn't notice this until months down the road. And I should mention we're on 2008 R2, but I think this works for both 2008 and R2. Just use a DISM command to upgrade in 2 minutes from STD to ENT:

1. From an admin cmd prompt, enter: dism /online /Set-Edition:ServerEnterprise /ProductKey:489J6-VHDMP-X63PK-3K798-CPX3Y

2. Allow the reboot, it will reboot once more after that.

3. From an admin cmd prompt, enter:

slmgr -ckms
slmgr -ato

slmgr -ckms clears the KMS entry and slmgr -ato activates a new one. What's the Product Key I'm listing? A generic, publically posted one you can use before step -ckms nukes it.

Oh, and this will require 2 reboots. No new patches will be needed either.

  • Created on .

Parent Site to Child Site Global State = Link Failed

CM2 (Primary site) was out of room (think it was the 9 million patches we downloaded).  It was probably like that for a few days.  It's a VM and yesterday I migrated CM4 (WSUS) off to another host so this box could get more disk.  That helped and things started purring, but software update summaries seem to be dead.  Earlier today I saw that DB replication was dead and it still is:  

Parent Site to Child Site Global State = Link Failed

Can't find boo on how to reinitialize it as it's service broker based, not transactional based.  Hoping it'll just catch up or something, but who knows?

Update: Beta 2 still has some transactional replication.  A couple of those replicas had gone stale.  I had to reinitialize them and generate a new snapshot.  That seems to be helping the console dates (one was status msgs).

Update: Found this in the primary (CM2) site's rcmctrl.log: ERROR: DRS queue is NOT enabled, so not calling the DrsActivation procedure.

So I opened SQL Server Studio and went to the service broker & enabled it.  Link status is back to active and data is flowing again!

  • Created on .

SQL issues in CM12

I submitted this bug on Connect the other day:

Both my CAS and primary have had this error repeat since install. No ill effects noticed, but I think we should know what causes it.
Cannot insert duplicate key row in object 'sys.syscommittab' with unique index 'si_xdes_id'
Failed to flush the commit table to disk in dbid 5 due to error 2601. Check the errorlog for more information.

They recommended that I install CU14 for old SQL 2008 SP1.  They also state that this won't fix the issue, but it supposedly will keep the boxes from having the issue after we make the fix.

So I downloaded CU14 and tried to install it 500 times before it clicked in my head that just possibly the download was corrupted.  Redownloaded it and it just installed on the primary fine.  CAS up next.

  • Created on .

CM12 - Day 53 in the lab

One thing I've been fighting is that some software updates aren't downloading. I'm working on that issue today. Some deploy, others have their last state saying they are downloading. I see that a folder is created in the cache, but no content populates it. There are a few known reasons for this in CM07 so I'll look into that today.

 

Also submitted a bug on SQL. I've been getting these repeatedly in the SQL logs:

Cannot insert duplicate key row in object 'sys.syscommittab' with unique index 'si_xdes_id'.

Failed to flush the commit table to disk in dbid 5 due to error 2601. Check the errorlog for more information.

Both the CAS and the primary show those errors since the day I installed CM.

  • Created on .

CM12 - Day 52 in the lab

Error sending DAV request. HTTP code 404, status 'Not Found'
GetDirectoryList_HTTP('http://mydp.mycompany.com:80/SMS_DP_SMSPKGD$/f9ff6d98-1a01-437e-8728-f29cb8c71b13') failed with code 0x8004027e.
Non-recoverable error retrieving manifest (0x8004027e).

Oops. Old boundary group had only a single subnet in it. So I just blew the other groups away and made one that included my single massive range of 10.0.0.0-255.255.255.254. See if that helps.

But that 404 is really saying content not found and the DP$ really has no files in it. Nor should it, right? Isn't that old technology? Don't we now pull from the content library instead?

Still have one SQL file trying to go to a dead DP. Guess I get to file a bug.

Just tried to install the software portal (can't call it shopping since 1E owns that name). The install failed. Investigating.

Fatal MSI Error - portlweb.msi could not be installed.

The WCF is not activated.
Solution: Make sure WCF is activated.

What the heck? Guess I should go read the prereqs. OK, installing the feature which also forces me to install .NET Framework 3.5.1. Something else I get to keep patching. Why can't .NET 4 just do this? Anyway, that worked.

  • Created on .
Copyright © 2021 - The Twin Cities Systems Management User Group