Monthly Archives: March 2014

How to reset a Cisco Catalyst Express 500 switch

Ever find yourself needing to reset a Cisco Catalyst Express 500 switch to factory defaults? Even though they’re pretty bad devices and have no CLI? Yeah, me too. I needed one for it’s PoE ports.

Cisco has a support doc here walking you through the process: http://www.cisco.com/c/en/us/support/docs/switches/catalyst-express-500-series-switches/70874-ce500-factory.html
If you have access to the CE 500’s Device Manager, then you’re golden. Just use the software reset via the web interface.

If you DON’T have access to the Device Manager, read on.

As of this writing, that doc was last updated in June of 2008, and in the years that passed since, Cisco failed to update it with one important detail. If you don’t already have access to the Device Manager, You must use a Windows XP machine.

Apparently, this is due to DHCP implementation on Windows 7. On a Windows 7 machine, you’ll never get past the DHCP Discover phase in the DHCP process.

After capturing packets in Wireshark on the Windows 7 client, after the client sends a DHCP Discover packet, the Catalyst doesn’t respond with a DHCP Offer. It appeares that the CE 500 doesn’t like Windows 7’s DHCP Discover packet,  and so it never responds.

On XP, it works just fine though. Take care when noting your IP address, I got a 169.254.0.2 address via DHCP from the switch, not the 10.0.0.2 as in the Cisco doc. At first, this made me think that IP was APIPA from Windows XP, and I started all over again. On the second go-round, I noticed the default gateway of 169.254.0.1, which of course does not happen with APIPA. After that, it was all textbook.

Setting up a SANless Windows 2012 Failover Cluster for SQL 2012 AlwaysOn cluster

Recently, our Windows 2012 SQL environment, running on Windows 2012 needed to move  to Windows 2012 R2, and carry the SQL Cluster along with it. Unfortunately, there was no upgrade path for this kind of move, so we decided to backup the databases, blow everything away, and start from scratch.

Additionally, we were using a SANless setup. Without a SAN, we were using SIOS SteelEye DataKeeper to provide SAN functionality. It was mirroring a volume on our Primary SQL server to the Secondary SQL server. And on top of this we were running SQL Server 2012 as an AlwaysOn availability group.

First, we backup the databases. Next, we made a careful note of the SQL server network name (the name of the AlwaysOn cluster), so we could use the name again, and the SQL connection strings wouldn’t break.

At this point, we’re ready to re-install the OS. After this is done, we go into Active Directory Users & Computers, and delete the old Failover Cluster and SQL Cluster computer objects. If you don’t, the new clusters will try to use the old objects, but the GUID’s wont match, and you’ll end up with cluster setup problems. Also, be sure to delete the old DNS records for these as well.

For the next step, follow these two walkthroughs. They’re excellent articles to take you through the rest of the process. One thing to note before beginning. SIOS DataKeeper introduced support for Windows Server 2012 R2 in version 8.0. Version 7.x and below only support up Windows Server 2003-2012, not 2012 R2. Make sure you’ve got the right version.

Clustering SQL Server 2012 on Windows Server 2012 Step-by-Step, Part I

Clustering SQL Server 2012 on Windows Server 2012 Step-by-Step, Part II

When you’re all done, hand it over to the DBA’s and the WEB guys and go drink a well-earned beer (or two).