We have a serial connection into the service processor and have done that.
When you do a start /SP/console - The system does not prompt or do anything, it is dead.
So we have done a reset /SP and even unplugged and restarted the system.
It comes up, puts out the BIOS messages and boots. You see it hitting the disk
and running. The after about 15 minutes it hangs again. No disk activity. You can
ping network interfaces but they give connection refused. The serial console stops
We have worked a bit with Oracle support to boot it with milestone=none at it will
come up and work to that point. However as soon as you do a svcadm milestone single-user it will
hang in about 15 minutes.
I suspect it is trying to restart the scrub on the storage pool it was doing but I do not
know how to unschedule that and have not see a scrub running before it dies.
Oracle suggested leaving it up all night, stating they see zfs_destorys occuring and that
it will finish and give us the # prompt back. However we have left it twice running overnight
and nothing is happening.
If your 7000 have 100% full for shares.. (file system), what about try to delete files like these...
#1. ssh 7000-ipaddress.
#2. login root id
#3. 7000>confirm shell
#4. 7000#cd exports/<100% full file system>
#5. move the location directory able to delete files.
#6. 7000#exports/volume1/aaa/rm <to delete files> <<- to delete files to file system usage going to 99%.
#7. then you can delete the files via CIFS or NFS...