[nodejs] Backport request for round robin scheduling in cluster

2013-07-13 Thread rdoherty
Hi, my current project is being bitten by the cluster module's reliance on the OS kernel to do balancing between nodejs processes. Out of 8 processes, 2 get > 70% combined requests/workload. This causes our servers to eventually stop responding, requiring a restart. I have tried to patch 0.10.1

Re: [nodejs] Backport request for round robin scheduling in cluster

2013-07-15 Thread rdoherty
Thanks! I will try that now. On Saturday, July 13, 2013 3:52:47 AM UTC-7, Ben Noordhuis wrote: > > On Sat, Jul 13, 2013 at 4:18 AM, > > wrote: > > Hi, my current project is being bitten by the cluster module's reliance > on > > the OS kernel to do balancing between nodejs processes. Out of 8

Re: [nodejs] Backport request for round robin scheduling in cluster

2013-07-15 Thread rdoherty
::HandleScope&’ ../deps/v8/include/v8.h:448:3: note: v8::HandleScope::HandleScope() ../deps/v8/include/v8.h:448:3: note: candidate expects 0 arguments, 1 provided make[1]: *** [/home/rdoherty/node/out/Release/obj.target/node/src/node.o] Error 1 On Monday, July 15, 2013 10:31:09 AM UTC-7, rd