Ah, the perils of auto-upgrades. Last night, Fedora Core 3 upgraded my
perfectly working nut installation to 2.0.2. I was informed of that
because every xterm in the house (about 8 over a Linux box, two laptops
and a G4 server) started beeping at 3:20am and didn't stop until I shut
down nut.
Keep in mind this configuration has been working for about two years.
Here's the messages I get when I attempt to start everything up:
Oct 7 07:42:59 allhats mge-utalk: updateinfo: Cannot update system status
Oct 7 07:43:04 allhats mge-utalk[1605]: Startup successful
Oct 7 07:43:04 allhats ups: mge-utalk startup succeeded
Oct 7 07:43:04 allhats upsd[1608]: Connected to UPS [evolution]:
mge-utalk-ttyS0
Oct 7 07:43:04 allhats upsd: Connected to UPS [evolution]: mge-utalk-ttyS0
Oct 7 07:43:04 allhats upsd: Network UPS Tools upsd 2.0.2
Oct 7 07:43:04 allhats upsd: Synchronizing...
Oct 7 07:43:05 allhats upsd: .
Oct 7 07:43:06 allhats upsd: .
Oct 7 07:43:06 allhats mge-utalk[1605]: updateinfo: Cannot update system status
Oct 7 07:43:07 allhats upsd: .
Oct 7 07:43:08 allhats upsd: .
Oct 7 07:43:09 allhats upsd: . giving up
Oct 7 07:43:09 allhats upsd[1609]: Startup successful
Oct 7 07:43:09 allhats ups: upsd startup succeeded
Oct 7 07:43:09 allhats upsmon[1613]: Startup successful
Oct 7 07:43:09 allhats upsmon: UPS: evolution@localhost (master) (power value
1)
Oct 7 07:43:09 allhats upsmon: Using power down flag file /etc/killpower
Oct 7 07:43:09 allhats upsmon:
Oct 7 07:43:09 allhats upsd[1609]: Connection from 127.0.0.1
Oct 7 07:43:09 allhats upsd[1609]: Client monuser@127.0.0.1 logged into UPS
[evolution]
Oct 7 07:43:09 allhats upsmon[1614]: Poll UPS [evolution@localhost] failed -
Variable not supported by UPS
Oct 7 07:43:09 allhats upsmon[1614]: Communications with UPS
evolution@localhost lost
Oct 7 07:43:09 allhats ups: upsmon startup succeeded
Oct 7 07:43:09 allhats wall[1616]: wall: user nut broadcasted 1 lines (50
chars)
Oct 7 07:43:14 allhats upsmon[1614]: Poll UPS [evolution@localhost] failed -
Variable not supported by UPS
Oct 7 07:43:14 allhats upsmon[1614]: UPS evolution@localhost is unavailable
Oct 7 07:43:14 allhats wall[1621]: wall: user nut broadcasted 1 lines (40
chars)
Oct 7 07:43:14 allhats mge-utalk[1605]: updateinfo: Cannot update system status
Oct 7 07:43:19 allhats upsmon[1614]: Communications with UPS
evolution@localhost established
Oct 7 07:43:19 allhats wall[1624]: wall: user nut broadcasted 1 lines (57
chars)
Oct 7 07:43:22 allhats mge-utalk[1605]: updateinfo: Cannot update system status
Oct 7 07:43:22 allhats upsd[1609]: Data for UPS [evolution] is stale - check
driver
Oct 7 07:43:24 allhats upsmon[1614]: Poll UPS [evolution@localhost] failed -
Data stale
Oct 7 07:43:24 allhats upsmon[1614]: Communications with UPS
evolution@localhost lost
I suppose the first thing to solve is why I'm getting that "updateinfo:
Cannot update system status" first thing. I tried stracing it, and it
appears to happen right after it attempts to write something and read
something to /dev/ttyS0. But I read a bunch of other stuff from
/dev/ttyS0 before hand - if I start up mge-utalk on its own, it first
detects that it's a Evolution 1500
write(4, "?", 1) = 1
nanosleep({0, 0}, NULL) = 0
write(4, "\r", 1) = 1
nanosleep({0, 0}, NULL) = 0
write(4, "\n", 1) = 1
nanosleep({0, 0}, NULL) = 0
nanosleep({0, 1000000}, NULL) = 0
select(5, [4], NULL, NULL, {3, 0}) = 1 (in [4], left {2, 934000})
read(4, "13200\n\r", 64) = 7
write(1, ".Detected Evolution 1500 on /dev"..., 39.Detected Evolution
1500 on /dev/ttyS0
) = 39
and later
select(5, [4], NULL, NULL, {3, 0}) = 1 (in [4], left {3, 0})
read(4, "0", 64) = 1
select(5, [4], NULL, NULL, {3, 0}) = 1 (in [4], left {2, 971000})
read(4, "XXX 11X1", 64) = 8
select(5, [4], NULL, NULL, {3, 0}) = 1 (in [4], left {2, 968000})
read(4, "1000\n\r?\n", 64) = 8
write(2, "updateinfo: Cannot update system"..., 40updateinfo: Cannot
update system status
--
Paul Tomblin <ptomblin@xcski.com> http://xcski.com/blogs/pt/
"and by God I *KNOW* what this network is for, and you can't have
it."
- Russ Allbery draws a line in the sand for Usenet.
(http://www.eyrie.org/~eagle/writing/rant.html)