I want to gather suggestions on the preferred cli email program to be used by smartcoin. I'll go with the most requested.
I finished a new routine which monitors for GPU lockups (the kind common when pushing overclocking a bit too far)
All of my tests so far indicate that a full reboot is needed to recover from locked card conditions. Currently, the routine will monitor for this kind of hard lock condition, and if found it will Log and error, shut down smartcoin and fire off an email. I'm going to do some tests to see if its possible to revive a locked card, but until then this is about as elegant as I can come up with.
Any further suggestions on this?
Why would you kill smartcoin and kill the other working devices? Let's say I go to bed and sleep 7 hours. After 1 hour one GPU locks. Then you kill smartcoin and 6 hours idle time for 3 other GPUs. That doesn't make sense to me. The current state is even better, because when getting up I can easily detect the the lockup by looking at the stats and I wouldn't waste 6 hours x3 GPUs hashing time.
Email, wonderful. Shutdown smartcoin, very bad idea imo.
most people don't wanna checking rig daily when they wake up.
maybe not shutdown,
restart (email about restart)
with smartcoin autostart after restart. (email about recovery)
then if restart 5x in 1hour (too hot outside or something) let it hang and send email every x hours about lockup.
also emails about configurable events will be awsome:
like
temperature threshold
failover - profile name
fan threshold
stats every X minutes
Mhash/s threshold
etc.