Remove the number of attempts/seconds part of the
query since the failure doesn't apparently always
ultimately fail on the same number so the query
was missing some hits. There will be some redundancy
now but we should have expanded coverage.
Change-Id: Ife831891a4c76106241761d6557c38c20d96f885
Related-Bug: #1349617
This query is ugly, but it's checking on the ssh protocol
banner failure and the end of the retry loop to catch the
failures. It's ugly, but it's 100% fail and mainly on
grenade jobs (with an ironic job sprinkled in), check and
gate, 28 hits in 7 days.
Change-Id: I241d1a0b1877fee8d9f5c774f70b6e880d2126fb
Related-Bug: #1349617