Just a heads up on how things are going with Ice and general redundancy measures.  I’ve got cron jobs in place to distribute SSL certs across all of the physical machines so any one physical machine being alive will be adequate to get access to SSL certs.

I’ve got a 5th name server running as a virtual machine on my machine at home so even if the co-lo is down entirely incoming e-mail will not be lost.

I’ve got a cron job copying /home and mail spool to another machine so either machine being down, I will still have a way of making these things accessible.

I’m still working on ice.  It turned out both Asus memory boards I have had dead memory channels, one because pins are missing from the CPU socket but the other is a mystery.  I can not see any missing or bent pins, yet, the BIOS sees all the DIMMS even though the CPU does not so this still points to a socket problem, perhaps a bad solder connection or open trace.

So I ordered another board, but did not pay adequate attention to the description, it was a parts board so not usable.  The seller agreed to take it back if I paid shipping even though he had said no returns up front.  Thank God some decent people still exist.

Then I ordered another board, this one supposedly working, well it arrived and had a minor problem.  Someone broke a screw off for the nvme drive in one of the mounts.  I was unable to get it out of the mount so ordered a kit for $6 off of amazon that contains mounts and mounting screws and I will just replace it.  A pain as I have to take the board out again as I can not get to it from the back.

But still moving forward…