2400IPX - SYSTEM DOWN!!!

[phadobas]

Last night there was some work being done to the power for our building and the backup batteries also ran out. When the power got restored, the system came up but gives no dial tone. It's a 1-stack, 3-PIM 2400IPX with dual CPU and dual TSW system.

Error messages say 1-A - Both TSW Failure (Permanent).

The CMOD command shows both TSW system in make-busy state. SINZ didn't help and complete power-down and power-up didn't help. I swapped the 2 SW10 cards with each other, but that didn't help eaither.

I got the Dterms powering on, and they all appear to be normal, but no dial-tone; obviously.

It's early hours, so my dealer is not yet reachable. Soon will be, but I'm looking for any help/advice I can get.

 

[OMBD]

Either SW10 or PC36 failure.

 

[phadobas]

We got the system up. I don't know if it's becasue we just tried to turn it on for the nth time, or becasue we did it differently. This time, we kept the power switch for CPU 0 off, before applying power to the bottom shelf (processor rack). Soon after, the system came up and got dial-tone. This is where we stand right now.
We are verifying operation of different functions and then have to check what recent MAT changes were not backed up before we lost power. Then we will turn CPU 0 back on and see if we can switch to that. Take it from there...

 

[belevedere]

There is a proper procedure for a shut down or reboot.
Shut down, CPU, TSW shelf (multiple stacks), PIMS.
Reboot, turn on PIMS first, TSW shelf (if you have one) then CPU's last. The CPU's need to see the time division up.
In a controlled shut down perform back up's first.

 

[phadobas]

OK, here is more information on this.
If the system comes up on CPU 0, it works and with CMOD, you can switch the TSW system from 0 to 1 and they both work. When you boot the system on CPU 1, it says both TSW 0/1 systems are in make-busy, unusable and you can't use the system.
Also, if you boot up on CPU0, and everything works, and you turn on CPU 1, and then try to CMOD from CPU 0 to CPU 1, the system will again say (under the command of CPU 1) that TSW0/1 are both make-busy, unusable.
Then we tried to switch the CPU cables around, and the problem remained exactly the same: system works under CPU0 but doesn't work under CPU1.
While I possibly have a dead CPU1, would it worth a try to re-format the HDD of side 1, reload system software and programming data and try again on CPU 1? Or does the system keep those HDD-s on side 0 and 1 synced? Or any other ideas?

 

[belevedere]

When you switched cables, CPU 0 became CPU 1. Did the problem remain with the CPU or move?
If the problem remained with CPU 1 with the CPU 0 cables, then I would re-flash and reload the CF in CPU 1.
You could use the CFCS command to stop the CF in CPU 0, then using a CF reader and the NEC CF program to copy it then load it on to the other CF. When you plug the CF back in to CPU 0 make sure that you use CFCS again to re-start it. CFCS is like the windows PC CF removal tool usually located in the bottom right tray of a PC.