Prohashing Mining Pool Forums

There will be system downtime in about three hours, around 4:00pm EDT, to install the memory in the mining servers. We'll have 256GB in these servers, up from 64GB. It arrived earlier than expected, and should resolve all issues associated with mining server reboots.

Things are looking up, eh? Thanks for the info.

I would think that quadrupling the memory should help quite a bit

Thanks for everything!

Thank you for the notice !

Steve Sokolowski wrote: ↑Mon Jun 14, 2021 1:08 pm There will be system downtime in about three hours, around 4:00pm EDT, to install the memory in the mining servers. We'll have 256GB in these servers, up from 64GB. It arrived earlier than expected, and should resolve all issues associated with mining server reboots.

4x the RAM and a couple days early, oh yeah ! Good stuff

Well.....

unfortunately, it's probably just a band aid for now. It will now take 4 times longer for the memory leak to eat up the RAM.

The real solution will be to figure out what is causing the memory leak and then how to fix it. If they are lucky it will just be a coding problem on their side and not something where they would need some other vendor to fix a bug in their software.

It's definitely a step in the right direction though. Keep up the good work guys.

Marduke wrote: ↑Mon Jun 14, 2021 11:29 pm Well.....

unfortunately, it's probably just a band aid for now. It will now take 4 times longer for the memory leak to eat up the RAM.

The real solution will be to figure out what is causing the memory leak and then how to fix it. If they are lucky it will just be a coding problem on their side and not something where they would need some other vendor to fix a bug in their software.

It's definitely a step in the right direction though. Keep up the good work guys.

I think that we fixed the memory leak, but the problem is that there isn't enough memory for the mining servers to get all the data they need. As the data fills up over two hours, after which it would normally be deleted from memory, the servers crash because all eight of them can't fit in memory at the same time. That leads to rolling crashes where memory is freed up, and then the restarted server starts using more and more, and the longest-running is killed, and so on.

Once enough memory is available to hold two hours of data, then we expect that the memory usage will stop growing.

Thanks for the hard work fellas!