Problem is that we haven't found a block for a couple days. I've checked the pool and related infrastructure twice, everything is working 100% correct. Its just bad luck. 

My update was meant for those miners who still haven't updated to a v8 compatible miner. If your rig is working, then you don't have to do anything.

Update: Be advised, since both XMRig and XMR-Stak slightly updated their configuration, you should not only update the miner but generate a fresh config as well. You can use the the Setup-Generator on the pool site to assist you with that.

Of course not. But the situation has been remedied long ago.

I wouldn't say that it doesn't deserve those shares. It did what is was being told to do so by the pool, unaware that the work was in vain. Just like any other miner connected to the pool during the same time-frame. 

The miner (the app) doesn't know that it is calculating hashes against a stale block because it does not have any knowledge of the current blockchain state. It doesn't have to because that's the pool's job. A pool that not receives work updates from the blockchain in a timely manner, can have an army of miners connected to it producing shares until the end of time without ever finding a block. That's why problems in coin-daemons are so serious. This is the reason why the Ethereum and Ethereum Classic pools are down until the related bug is fixed.

That's not what I meant. Fortunately the scenario you've mentioned isn't even possible since we validate each and every share (some pools start to relax checking once a miner has submitted a couple of valid shares - we don't). What I was talking about is the coin specific blockchain-daemon powering the pool which in turn distributes work to the connected miners. If the blockchain-daemon isn't notifying the pool of new blocks (and this is what happened), then the pool is unable to feed new work to the miners. Does this explanation make sense to you? Otherwise I can go into more detail.

I've spent the entire weekend optimizing our monitoring kung-fu in order to detect and react to cases where a coin daemon stops publishing updates for a while.

Looking better now.

Done.