Hi Oriol,
I'm sorry for the inconvenience. I thought I had all the issues fixed. The root cause was a file descriptor leak in Jenkins. Every time someone visited one of the pages there, Jenkins would leak 2 descriptors. For some reason, we had a huge spike in overnight (eastern time) page hits starting March 7. I increased the open file limit, but the following night, the limit got breached again. I updated Jenkins to a version that no longer leaks file descriptors but now it seems Sonar got corrupted and crashed the application server. I've reinstalled Sonar. I'll check the logs again tonight to see if the problem still persists. It should be working for now.
Access logging was disabled by default so I've enabled it. That should tell us if we're getting hammered by search engines, hackers, or other bots.
Thanks,
Jason