Possible dashboard "query stampede" after restart

[nblumhardt]

Seq has some mitigations to ensure that the requests from multiple dashboards don't swamp the server upon restart, but it appears that particularly I/O intensive charts could still cause restart issues:

http://docs.getseq.net/v4/discuss/592dee37fe3c260f00bad695

Some additional checks might help here.

[nblumhardt]

I think #579 may have been the primary issue responsible for this one, but in 4.1 to even out resource usage the dashboard will update each chart serially in turn.