So I thought that when Claymore introduced the -minspeed variable, that was pretty cool!  It's kind of like a low hashrate watchdog, right??

So I set it, with a decent margin, so that if a GPU was way out of whack it would reboot and all is good.... but I'm telling the story backwards... let me rewind.


This morning, the DAG update happened.  Strangely, one of my rigs kept on rebooting.  Like, LOTS of reboots....Not being in the same room, with most feedback from EthMan, nothing was super obvious.  Each GPU's hashrate decreased just a bit as expected, but none of them dropping to zero, or anything typically suspicious.  I was thinking that MAYBE the DAG got heinous enough to be difficult with the strap timings.  This rig has RX 580 Pulses, and I haven't been successful getting tight straps on it (like, no Ubermix or whatever) but it's "ok" - but maybe this latest DAG was just too much.

I tried to downclock the memory... -30, -60, -100... nothing!  Maybe, I'm thinking, maybe I finally install that bobo driver and see if that helps? 

So I finally got to watch the console ... weird, no GPUs are misbehaving at ALL.  WTF?

FIVE MINUTES (this is a magic number) and the error appeared.  It pushed that one rig down under the -minspeed threshold.   That rig rebooted 73 times today.  SMH.

Anyhow, changed the -minspeed to a now-safer number.  Ta f'ing Da.