Is there a (documented) way to get several simulators into Operations Manager? It seems that due to the fact that mny sims all have the same serialno/SystemID so OM seems to think they all the same machines. I am trying to get a Protection Manager Demo running and need at least two 8.0(.1) sims for that...
I have found that at bootup the Sim looks in /sim/dev/ for a file called ,serialno (just like the 7.x Sims), but I cannot access that file, as the systemshell demo user does not have right to create that file...
The one thing I'd add is that if you make that serial number change in the SIMLOADER before first setting up the simulator, then it won't be necessary to go through the maintenance mode process to reassign the disks. The initial setup will use the serial number that's defined at the time.
Lastly, we are planning to add functionality to automatically generate unique serial numbers for simulators in a future release of Data ONTAP.
Just a quick addendum: I've recently found out that there's another boot variable that takes precedent in setting the UID's. If you're looking to avoid UID conflicts, the best thing to do is set bootarg.nvram.sysid and SYS_SERIAL_NUM to the same value, and do this prior to first boot and initializing the disks. That prevents having to go through maintenance mode to reassign the disks. Other UID's may also be fixed and harder to change after the first complete boot.
Also, the system ID format is a 10-digit number. The last two digits of the number need to be unique within the C-Mode cluster for the disk UID's to be unique. So one way to manage the values is to use the first 8 digits to represent the cluster, and the last two digits to represent the node. For example: 1111111101, 1111111102, ..., 2222222201, 2222222202, etc.
I tried this method on a fresh install of the sim, I drop to SIMLOADER and enter in the "set" commands listed, boot, hit ctrl-C to enter the boot menu, and select "4" to initialize the disks and wipe config. Then the VM dumps core and reboots without doing the config wipe / disk init.
I unzipped a completely fresh copy of the guest, so I really doubt its vm corruption. If I unzip the guest and DON'T do the simloader commands it works and initializes just fine.
Are you using one of the original 8.0.0 simulators? That method may not work on the older vsims. If this is happening with the DOT 8.0.1 vsim, then the only thing that comes to mind is to make sure that the above steps are performed as the first thing after the simulator is extracted from the zip file and powered on.
Take care and hope that helps,
p.s., I recently left NetApp and don't have access to the vsims right now to verify that the sysid fix panics the DOT 8.0.0 sim.
I downloaded vSIM 8.1, installed it on ESX4i, at first boot stopped in SIMLOADER and changed both SYS_SERIAL_NUM and bootarg.nvram.sysid (just one number down from original). I then continued with SIMLOADER> boot and stopped in special boot menu, selecting option 4 for a fresh install.
Now I've got two vSIM 8.1 in OnCommand/DFM with unique SN and sysid .
Yes, seems that happens in some DOT8.x simulator versions. In that case follow these steps to change the serial number (you'll need a fresh simulator to do this, can't use one that already ended up in the panic situation).
Set up the simulator, e.g. at first startup, Ctrl+C into Boot Menu and choose option 4 to initialize, then go through setup to assign hostname, IP, etc.
"priv set advanced"
"useradmin diaguser unlock"
"useradmin diaguser password" and enter a new password
"systemshell" and login with username diag and the password you just created
"sudo rm /sim/dev/,disks/,reservations"
Reboot VM, Ctrl+C at first prompt so you get into the Loader
"set bootarg.nvram.sysid=1111111101" --> change number as needed
"set SYS_SERIAL_NUM=1111111101" --> change number as needed