Glitch when selecting a large time range in ASAP::O-tv
We noticed (during the P11 experiment) that selecting a large time range in ASAP::O-tv, say 24 hours, causes the web service to temporarily stop responding. When it comes back, there's a gap in the graphs. See for instance the receiver cache memory usage graph here (for those with access to the P11 project chat).
Our hypothesis is that the large range causes some database system to OOM, and stop collecting monitoring data until it recovers.
Workaround: avoid selecting large time ranges.