Hashrate Watchdog - multiple miners

[matej.paciga]

Hey guys,
I run HiveOS with three different miners under one main worker (mixed setup due to some limitations etc.)
Wondered to ask - is there a way to monitor the overall hashrate or to monitor all three miners differently?

What I can see from the logs is that no matter what, the hasrhrate watchdog is only considering first miner and its values.
Change from miners to algos in hashrate watchdog are making no difference (only hashrate of first miner is considered).

Is this known behavior?
What should I check or modify to improve the behavior?

Many thanks!

[mthqwork]

I have the same issue. Using TeamRedMiner as first, and Gminer/Trex as second.
Hashrate watchdog does not care about other miners but first. This should be fixed, as i already had many times, that I noticed only long hours later that my rig only performing 50% of its hashrate.

[sonic6]

So for me. TRM and gminer:

[matej.paciga]

I’ve workaround it with “WD power” watchdog since in my specific use-case, there are 3 old R9 cards using lolminer which drops the hashrate and while doing so, the power consumption is lower as well.

My combination is to have:
Hashrate watchdog for T-rex
WD power watchdog for reported consumption of all GPU’s and minimal power threshold set to approx. 5% drop, which corresponds to 10%+ of hashrate drop in my case.

Works well

[matej.paciga]

Unfortunately, after some time I’ve found it doesn’t work as expected.
WD power watchdog is somehow stressing the whole HiveOS and I’m seeing strange behavior and even crash of Hive itself.
I’m disabling it for now and changing the order of miners to monitor “lolminer” for hashrate instead of Trex on which GPU’s are stable

[mthqwork]

One year passed, still an issue. Can’t believe that this use-case is such unimportant.

[painting]

I’m also seeing this problem. Set the algo watchdog for a resonable value for the total rig’s hashrate. However, the watchdog only takes notice of the first miner working on that algo. Probably the devs’ logic was that you would mine one thing on a miner and something else on the others.

Had the kawpow watchdog at 200 Mh/s, and since trm was only doing 25, it restarted the miners

I tried not setting a value, but hive has a fit about that