Hi all,
We have a vsphere infrastructure with 3 esxi 5.5 in HA and datastore on SAN.
Every host run from SD his own card and has his own SATA disk.
Every host is configured to put vmsyslogd files and dump files on his own SATA disk.
On every host SATA disk there is also configured a scratch partition.
On one host i run into a strange I/O error during a simple 'cat /var/log/hostd.log'.
The host seems to run just fine: no disk errors, no strange things happens.
I make a request to vmware support to investigate that error, and they found a problem on the SATA disk.
They found SCSI related read errors on vmkernel.log, so they asks me to contact my disk vendor to change the faulted disk.
Before changing the disk i think i need to change vmsyslogd and dump file location and create another scratch partition so the esxi can run as usual.
Can i temporary move all things to an USB disk so i can safely change the SATA disk?
I don't want to mess with SAN and SD card so i'm looking for the most simple and safe way to do that...
Thanks and sorry for my poor english :-)