RPS failure in 5308xl

[Les Ligetfalvy]

I am testing a bunch of these switches before putting them into production and thought I would share my experience.
The #1 power supply (I also have the 2nd RPS) was showing a brief fault at powerup but the fault would clear after boot. There is a shortcoming in the OS that prevents boot time failures from reaching the trap receiver (HP is working on that issue) so there is no reason to suspect that there is something wrong (remember, the fault light is off). Removing power from the #2 (good) RPS causes the switch to reboot over and over... This may be freak occurrance but if the fault goes unnoticed... well you see what I mean.

I designed the power source for these switches to come from separate distribution points so that if the power loss is isolated to just one, the switch stays up. Knowing that, it is highly likely that the two sources may not both come up at the same time, so a lack of power indication in the boot log could be easily dismissed and since the fault does not persist for long after boot, there is no reason to suspect the RPS.

[OLARU Dan]

Les,
If you swap #1 and #2 what happens? If the problem moves to #2, then you have a faulty power supply module. If not, the fault is in the RPS bay or beyond, inside the switch.

[Les Ligetfalvy]

Dan,
Thanks for the response. I apologize for failing to mention that I already did complete troubleshooting and that HP replaced the RPS. I posted this more as a heads-up that the extinguished fault light cannot necessarily be trusted.

You are not going to believe this... the replacement power supply I just received is churning out hundreds of errors in the event log. Bear in mind that when the original RPS failed, I took power supply #1 from 2.23 and swapped it with power supply #1 of 2.21 and the problem moved with it. Later, I replaced it with my own in-house spare (2.27) and that went well. When I got the replacement from HP, it did not go into 2.23 where the original failed nor did it go in 2.21 but rather it went into my spare chassis 2.27 from which I borrowed the RPS.

I ran through several cold boots powering them up in different sequences but the errors are consistent. This RPS must have been a problem return that escaped rigorous testing. I am now waiting for HP to replace the replacement.