Quote from: smoothrunnings on February 23, 2014, 03:16:10 AM

Quote from: oldbushie on February 23, 2014, 12:56:19 AM

Time for a reboot/purge/reinstall?

I noticed that somehow 2 p2pool instances were running on my VPS for the past week (I think one of the auto-check scripts flubbed) so I just rebooted the VPS. Though my shares found rate was still ok considering...

VPS?

A VPS is a virtual private server -- roughly equivalent to a shared hosting account, but emulating a dedicated server with root access (or equivalent) instead of just a Web site. Typically these offers come with significant "fair use" restrictions to make sure no one customer totally overloads or monopolizes the physical resources, starving the other VPS's hosted on that machine.

I've actually been having the opposite problem from oldbushie. On my VPS, I've installed a monitoring utility (monit). For some reason, monit determines that p2pool and bitcoind aren't running (the email explicitly says the trigger is that the process isn't running, not that one of the other trigger conditions is met), about once an hour, and promptly proceeds to kill and restart the existing sessions. This is surprising, since both ran just fine for days before I installed monit. My guess is I configured monit wrong. If anyone has experience debugging monit scripts or settings, I'd appreciate any feedback.

Code:

check process bitcoind with pidfile "/home/USER/.bitcoin/bitcoind.pid"
start program "/usr/bin/bitcoind -datadir=/home/USER/.bitcoin -daemon"
as uid USER and gid USER
stop program "/usr/bin/pkill bitcoind"
as uid USER and gid USER
if failed port 8332 for 3 cycles then restart
if cpu > 60% for 4 cycles then restart
if cpu > 80% for 2 cycles then alert
if 2 restarts within 10 cycles then alert

check process p2pool
matching "p2pool"
start program = "/usr/bin/python /home/USER/p2pool/run_p2pool.py --address 1F2jgm72pw7VH9GKFMgHTqqUhxXSptADkV --fee 1 --give-author 1"
as uid USER and gid USER
stop program = "/usr/bin/pkill p2pool"
as uid USER and gid USER
if failed port 9332 for 3 cycles then restart
if cpu > 60% for 2 cycles then alert
if cpu > 80% for 10 cycles then restart
if 2 restarts within 10 cycles then alert