AD should not flood Jenkins

[astellingwerf]

When Jenkins is down, it seems as if scan jobs are queuing up. When Jenkins gets back up, I often need to block AD from accessing it, as it will hang Jenkins. (Now, to be honest, we have complex views; composite, containing long chains of include and exclude filters.)

My suggestion should be:

• Don't spawn a new scan job before the previous one is done. Group by ???
• Put a timeout on the scan job
• If a timeout occurs for the 2nd or 3rd time in a row, switch from GETting to HEADing.
• If a HEAD starts to return HTTP 200, wait for some more time (maybe a few minutes, Jenkins start up takes quite a while) before actually issuing the heavier GET requests.

(I did not look into the AD code, so I might be jumping to conclusions...)