Slight jump in invalid shares this morning with Strix RTX3090

[Eon046]

Hi,

I have noticed a jump in invalid shares on my 2 GPUs rig. Usually 0.50 %, today: 2.30 %

I’m running 2 Strix RTX3090 O24G
They are sitting around 39 et 45°C core temp (no memory probe), fans at 85%, pulling 118 mh/s each.

Memory 2000 mhz, core 1100Mhz, they are currently drinking between 304 and 307w each.

I don’t believe this is quite an aggressive OC settings, is it?

Miner’s uptime is 6days,
mining on Hiveon (eu)
Ethminer v0.19.0.2
HiveOS: 0.6-210@210925
nVidia driver 460.84 (early summer 2021)

I assume ethminer v0.19.0.2 is current.

Are there incentives in getting my nVidia driver updated? Is there be such a thing as recommended driver version?

Please, let me know if you have noticed such a jump in invalid shares as well.

Thank you for reading, happy mining!

[ForgottenNavigator]

Do you mine to Hiveon?

My rigs are going nuts this morning as well. My stale shares are 4x their usual and where I was getting 1-4 invalids per day and now I have over 100 already today. All of my rigs are throwing the same number of invalids, where before one of my older rigs was the culprit.

When I checked pings to the pools I noticed my pings way higher than previously. Switched my rigs over to another network, same issue.

[Grea]

If the invalids suspicion is Pool server, might be worth pointing at another location even if the stales creep up.

US-East does not seem to have any issues and these are coming from (2) different ISPs, locations, etc:

[ForgottenNavigator]

Thanks, I’ll point to US East and see what happens. 143 invalids now. This is nuts.

Update: Swapping to US East seems to have fixed the majority of my issues. Guess I was too busy thinking it was an issue on my end. Thanks Grea.

[Eon046]

I’m mining on Hiveon indeed, @ForgottenNavigator , see stats here.

Yesterday, I lowered OC (mem from 2000 to 1900, core from 1100 to 1030), changed the pool url to the US East ones and my invalid shares rate hasn’t moved: I’m around 2.28 % still.

I have lowered them even more today. I’ll see what happens, eventually switching to Canada URLs later to day, stay tuned

Edit: After 4 hours with 1000 mhz core and 1800mhs memory, I dropped to 1,70% invalid shares on US East. I’m now going to shift back to EU hiveon pool adresses.