Unsolved
This post is more than 5 years old
12 Posts
0
1428
PE 1800 power drops on one ps for 10 seconds.....what?
Good morning, all!
A most interesting conundrum just came up in my data center. There are a few PE 1800 servers that have one power supply dropping off the bus for ten seconds, then returning. It happens infrequently at various times, and the other redundant p/s always takes the load.
The oddity is the ten second gap. I haven't caught the server in failure, and the p/s always comes back online.
Any thoughts on this? Is this normal behavior for a slightly rickety p/s, and will it fail hard one of these days?
Thanks in advance!
Gregg
DELL-Daniel My
Moderator
Moderator
•
6.2K Posts
0
August 7th, 2012 13:00
Hello Gregg
This could be a BMC or BIOS issue. It could also be a faulty power supply or power distribution board. Where are you seeing this failure take place? Is it hardware log errors indicating failure and then the PSU coming back online?
I would recommend making sure the system BIOS and BMC/ESM are at the most current revisions. Also, if there are firmware updates for the PSU's I would apply them. If the error persists then swap the PSU positions. If the error follows the PSU then the PSU is likely failing. If the error stays with the slot then the PDB is likely failing.
Thanks
DELL-Daniel My
Moderator
Moderator
•
6.2K Posts
0
August 7th, 2012 14:00
I was referring to swapping the Power Supply Units(PSU) not the Uninterruptible Power Supplies(UPS), but it is good to know that the PSU's having problems are not connected to the same UPS.
A BIOS or BMC update is not likely to resolve the issue if the amber light is on the PSU, but I would do it anyway. I would still recommend following the procedures I outlined.
Thanks
gregg2238
12 Posts
0
August 7th, 2012 14:00
Hi, Daniel!
Thanks for the quick response.
After getting a quick maintenance window I got a look at the SEL and found that a second PE 1800 was having the same symptom. They're cross-connected, so the PS2's are plugged in to different UPS circuits. I have yet another 1800 that's flashing amber, so it's quite possible that we have an epidemic.
In that case, would a BIOS update be more likely? I've seen various components get weird seemingly all on their own a few times; that'd be my worser-case scenario.
Thanks!
Gregg