asking for and not receiving a zone file isn't necessarily a problem, since the "node" could be a load-balancer

This commit is contained in:
Jude Nelson
2018-04-13 19:14:49 -04:00
parent ab519b4c8a
commit 8a6264508a

View File

@@ -3407,7 +3407,7 @@ class AtlasZonefileCrawler( threading.Thread ):
with AtlasPeerTableLocked() as ptbl:
# if the node didn't actually have these zonefiles, then
# update their inventories so we don't ask for them again.
# TODO: ban nodes that repeatedly lie to us
# TODO: do NOT ban nodes that repeatedly lie to us, since the "node" could be a load-balancer for a set of nodes that may or may not have the zonefile
for zfh in peer_zonefile_hashes:
log.debug("%s: %s did not have %s" % (self.hostport, peer_hostport, zfh))
atlas_peer_set_zonefile_status( peer_hostport, zfh, False, zonefile_bits=missing_zfinfo[zfh]['indexes'], peer_table=ptbl )