Work Restart Errors (Just started) [SOLVED]

Hey Guys;

I have 4 ASIC units working on the pool. I just added the 4th today, and noticed that when I did so, one of my other workers suddenly (and this had not happened yet) started flashing the red error saying that the miner was not responding to work restart testing.

Since these are all Zeus/Zeus clone miners, I'm happy to add a r=0.35 (which is what I saw in a different forum post for Zeusminers) to the passwords, but I do find it strange that it just started suddenly.

In the interests of full transparency, before I had set up a separate worker for the 4th ASIC unit I did add it as an additional device to another BFGMiner controller (I had ZUS 0 and 1) and after I had added it initially it seems the errors started. The configuration for all 4 units is now standalone controllers, so the 0/1 configuration has been eliminated.

Any hints?

I'll check with Chris to confirm, but my guess is that something happened to that miner, which caused it to take longer than 60s to spin up. It might have something to do with hardware errors (like "invalid nonce.") Let me see what Chris says and I'll get back to you later in the day.

The only errors I've noted in the miner output is just old block errors, which the reset of the scanning to 1s fixed for the most part. I'll see if I can up the verbosity of the output and look more closely.

It is odd that it just started happening as I tied the two units into the single controller.

Oh, also worth noting. Each miner and controller has a separate worker, so adding the second unit under the same worker name might have resulted in some database inconsistency? How about just giving the worker/miner that is in the red a new worker name?

I'm going to try this work restart delay for my TITANS will this lower the hashrate?
d=8192 n=titan_1 r=0.35

Pleased to report that after the system maintenance yesterday and my miners restarting their daemons, the errors have gone away.

For whatever it's worth, I've found that the best difficulty for Titans is 16384, and that just bouncing the miner process is enough to get the restart testing issue resolved. If you really want a static restart delay, then .8 is about where they average out.

As another follow up for fellow Zeus operators, I updated my BFGMiner software on all my controllers to 5.4. Some of the 5.2 versions seemed to be inconsistent and causing some problems too.

In short, I use d=2048 r=0.35 and BFGMiner 5.4 and things are now running more smoothly (with a slightly higher hashrate) than before with no restart problems.