After 9 days of frustrating work to replace our failing main web server, we’ve implemented plan K (plans A – J failed!). Plan K involved making a previously unreliable server (lingling) reliable again by installing software RAID and then moving all of the services onto this machine. We just finished with that and the new web server seems to be handling the load well so far.
Now its up to Dell to fix our server — fortunately, much to my surprise, we still have a service contract on that machine. So, its up to Dell to diagnose and fix the box and then we’ll put it back into rotation. Also, Steve Wyles (inhouseuk) has agreed to lend a hand in diagnosing the issues on moose, the new Sun Server — we’re suspecting a copy of Solaris that is too old to handle this new architecture.
So, expect more downtime in the future as we shuffle fixed machines back into rotation. Hopefully we’ll be able to proceed with a more sane pace from here on out.
Once again, giant kudos to Dave Evans for all his hard work! Thanks Dave!