Comment 6 for bug 1655608

Revision history for this message
clayg (clay-gerrard) wrote :

the best idea for now is to add some code around where we log that message we would go check handoffs and quarantine if the # of 200's was under some threshold - maybe 1/2 data or 2x parity - like if you have TOO many missing fragments quarantine is safe enough to assume the fragment was dark data instead of hoping "someday" the cluster will find enough fragments to rebuild the object to full health.