Hi all, As an FYI: the Jenkins master got an upgrade to a much beefier machine over the weekend. Per Infra (on bui...@apache.org): > Jenkins has been migrated to a brand new 48 core machine with 128GB RAM and 3.5TB of NVMe storage. Initial testing looks good, and builds are proceeding as expected. We hope this will provide a significant improvement in user experience on the front-end, and increase the throughput of builds. > > Please note that the 3.5TB of NVMe is less than what the previous jenkins master had. It is critical that all projects aggressively trim their retained builds in order for this shared resource to continue to function optimally. Anecdotally, the Jenkins UI feels much faster to me at this point, which is exciting! If anyone sees issues with e.g. scripts, access, builds queueing, or other Jenkins-related things please ping on this thread and we can put together a set of escalations to Infra.
As I recall we are not one of the projects causing problems with retained builds (I think we generally keep artifacts for <1mo) but it's worth noting the paragraph about storage space on Jenkins. If storage constraints give Infra problems they may implement a global retention window which would affect us at that point. Best, Jason -- ------- Jason Kuster Apache Beam / Google Cloud Dataflow See something? Say something. go/jasonkuster-feedback
