I'm not sure if this could be categorized as a defect, but I'll try anyway.

On my current project we're running Hudson (1.293) with no slaves defined. There
are about 150 jobs configured. All of them are configured to poll the subversion
repository every 10 min. The subversion repository is located on a different
server than the Hudson instance.
When Hudson is starting the cpu load on the server where subversion is running is
increasing dramatically and it takes about 40 minutes to boot Hudson.
Could this be a misconfiguration or is this normal behavior for Hudson running
that amount of jobs?