Solaris is not very good when it comes to dealing with broken hardware. Sure there is the fault management framework, but that does not mean that the problem is always detected by it. Instead sometimes you have a dead or dying disk drive that spams syslog with annoying messages until you physically unplug the drive from the system. This is what happened to me recently with our live storage system which is based on Solaris.

Thankfully, pulling the drive did fix the error messages from appearing, but then I had to do a format of the disk to make sure that it was OK.

The basis of the Solaris kernel has some really well tested code, but it seems that a misbehaving driver can cause all sorts of problems. This is also seen when a disk is dying and you do a zpool status and it takes forever for the info to come up. It’s a shame that is the state of things right now.