View Single Post
Old 05-07-2013, 06:20 PM   #36030
poohbear_nc
Now what?
poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.poohbear_nc ought to be getting tired of karma fortunes by now.
 
poohbear_nc's Avatar
 
Posts: 58,871
Karma: 135181808
Join Date: Feb 2009
Location: Durham, NC
Device: Every Kindle Ever Made & To Be Made!
So -

1. Our company uses thin clients at employee workstations.
2. All are plugged into surge protectors. The server is plugged into a UPS - the server is running Fedora Linux.
3. We had a really bad thunderstorm yesterday - with very close cloud to ground lightning strikes.
4. After a close strike, the power went out for about 20 seconds, then came back on.
5. The server was unaffected.
6. All of the thin clients lost connection to the server - this is normal when we have a total power outage. Normally we just power them off & on - and everything is fine.
7. Half the thin clients remained unable to boot up - even after repeated power cycling.
8. Thinking that we had a power surge when the electricity returned, we theorized that some of the thin clients had blown an internal fuse.
9. We opened them up - there were no fuses. No visible damage. Fans & connections tested normal.
10. So we cleaned them out (since they were open) and kept trying to figure out why some of the thin clients were seemingly dead.
11. My assistant pointed out that they may have over-heated since the fans weren't running during the outage. This sounded a bit extreme for such a small outage, but I said OK - let's hook one of the "dead" ones back up to the server.

PRESTO - the thin client booted up and acted as if nothing had happened.

12. Thinking it was a heat issue, we allowed the remaining "dead" thin clients to cool down. But after 20 minutes, they were still "dead" - unable to boot up.
13. Opening the "dead" units was the only way to restore them to a bootable condition. Weird?

14. Our current theory is that some of the units (who knows why) built up an internal electrical charge during the power surge - and only opening the metal cases and touching the internal components drained the charge and allowed the unit to boot up again.

We almost boxed all them up to ship off for service!

A. Anyone in IT every experience a situation like this?
B. If so, any explanation for how/why opening the case restored the units to functioning?

Or should I start making regular sacrifices to the deities of IT mystery?

We're still scratching our heads over this one!
poohbear_nc is offline   Reply With Quote