forums.parabuildci.org

Three other contributing factors: 1. We are using Nagios to monitor devices on our network. This periodically checks the 8080/tcp port; depending on the polling interval, it will lead to an increase of connections over time. To help deal with this, change the TCP_FIN_WAIT value on Linux from 60 to 5...

I am posting this thread so that there is a record of how I went about solving this problem. It might be unique to our environment, but others might benefit from my experience. This weekend, we rebooted many servers in our lab, including our Parabuild master server and build farms. After rebooting t...

Thanks for the update! That's a lot of changes. What do you thing has contributed to the result? How many virtual CPUs did it have before? Slava I suspect that the increase from 2 vCPUS to 4 vCPUS made the biggest difference in performance here. Between HTTP serving, build scheduling and handling H...

We are running the Build Manager in a virtual environment. Base hardware: ⋅  Dell PowerEdge 1955 ⋅  2-socket, dual-core 5130 CPUs @ 1.994GHz ⋅  HyperThreading Disabled ⋅  VMware ESXi 4.1.0 build 260247 ⋅  2 VMs running on the system ⋅  VMw...

We're seeing this CPU exhaustion issue in our environment as well. Build Manager Environment ⋅  8GB RAM ⋅  2 vCPU ⋅  Red Hat Enterprise Linux 5.5 64-bit ⋅  Oracle/Sun JRE 1.6.0_23 64-bit ⋅  Parabuild 4.0.55 ⋅  JAVA_OPTS="-Xms4096m -Xmx...