Topic: Cluster mining limitation? (Read 1733 times)

Okay, throwing a question out here. Please forgive me if this is not the correct place to post it. I am using a wallet customized for BitQuark, (a Quark/Bitcoin hybrid, wallet v0.8.3.7-g-beta). It may or may not pertain to the native Bitcoin wallet but I'd like some feedback from other wallet users anyway.

Over the last several days I've been tinkering with cluster mining (2 or more computers mining on a local network all pointed to the IP of the PC with the wallet being mined into). PC1 has the wallet and runs sph-sgminer 4.1.0. Other PCs on the network run minerd pointed at PC1. There seems to be a limit of 4 total mining clients to the wallet server daemon,

loop-back of sgminer on PC1 plus 3 other computers running minerd pointed at PC1, or
look-pack of sgminer on PC1 plus loop-back of minerd on PC1 plus 2 other computers running minerd pointed at PC1.

Any time I exceed that limit the additional computers return json_rpc_call failed. If I release one of the running clients the one that failed immediately takes over. However, even 4 total seems to be pushing it as it sometimes causes sgminer to sit and wait for work from pool 0. Three wallet mining clients seems to be the max stable limit. It also does not matter which PC on the network I make the wallet server. Everything from a Asus Atom 525 netbook up to a Core 2 Quad server with Raid 10. They all behave the same.

Is this a known limitation? Unique to BitQuark? Fixed in later versions of the Bitcoin wallet? Should I just be mining into a wallet on each PC anyway?

Z