Here's a short report from the administrator of OnEarth, Lucian.
While it was pretty inconvinient for this problem to occur, it's reassuring to
know OnEarth _is_ being looked after.
Cheers, Mikel
..............
There was a powerdown on Friday. My server was not supposed to go
down, but the conditioner tripped. It only came back on Monday,
and the server was up shortly afterwards. Then I had to catch up
with the daily MODISs, which kept crashing for a few days. Turns
out there is a new bug in an updated system library that the
reboot activated, which made the image handling code exceed 1GB
and crash. This affected the main server, but only for
large/complicated requests, so it crashed and restarted at random
times.
Took me a while to figure out the problem and a workaround, and I
was lucky!
It should all work fine now, but I said that many times before!
Of course, I didn't get around to update my caching server to
handle the 12 months of BMNG, or the 500+ days of MODIS.
|
