[Varnish] #834: Varnish sends health checks to backends after backend was removed with a config reload

[Varnish]

#834: Varnish sends health checks to backends after backend was removed with a
config reload
--------------------------------+-------------------------------------------
 Reporter:  mkania              |        Type:  defect
   Status:  new                 |    Priority:  normal
Milestone:                      |   Component:  build 
  Version:  2.1.3               |    Severity:  normal
 Keywords:  healthcheck reload  |  
--------------------------------+-------------------------------------------
 I noticed this on a production cluster, where we were removing backends
 from the varnish config, and issued a config reload. Requests no longer go
 to the removed backends, but varnish is still sending health checks.

     0 Backend_health - asset_demux_phx5 Still sick ------- 0 3 5 0.000000
 0.251935
     0 Backend_health - asset_demux_phx12 Still sick ------- 0 3 5 0.000000
 0.261331
     0 Backend_health - asset_demux_phx6 Still sick ------- 0 3 5 0.000000
 0.267248
     0 Backend_health - asset_demux_phx13 Still sick ------- 0 3 5 0.000000
 0.267399
     0 Backend_health - asset_demux_phx8 Still sick ------- 0 3 5 0.000000
 0.259694
     0 Backend_health - asset_demux_phx0 Still sick ------- 0 3 5 0.000000
 0.267166
     0 Backend_health - asset_demux_phx2 Still sick ------- 0 3 5 0.000000
 0.274461

 I've verified that all references have been removed from the config.

-- 
Ticket URL: <http://www.varnish-cache.org/trac/ticket/834>
Varnish <http://varnish-cache.org/>
The Varnish HTTP Accelerator