My exit
node's consensus weight just jumped from 20 to 1750
overnight. When I checked to see how things were going, my
log file is full of nameserver problems, happening every
couple of minutes:
Jan 31 14:12:40.000 [warn] eventdns: All nameservers have
failed
Jan 31 14:12:40.000 [notice] eventdns: Nameserver
8.8.4.4:53 is back up
Jan 31 14:18:35.000 [warn] eventdns: All nameservers have
failed
Jan 31 14:18:35.000 [notice] eventdns: Nameserver
8.8.4.4:53 is back up
Jan 31 14:20:53.000 [warn] eventdns: All nameservers have
failed
Jan 31 14:20:53.000 [notice] eventdns: Nameserver
8.8.4.4:53 is back up
Jan 31 14:20:59.000 [warn] eventdns: All nameservers have
failed
Jan 31 14:20:59.000 [notice] eventdns: Nameserver
8.8.4.4:53 is back up
But the "All nameservers have failed" and "Nameserver xxx
is back up" messages happen in pairs at the
exact same time. What's going on here, and is there
a way to fix this? My VPS has 2 nameservers listed for it,
should I be using those?
The times in tor logs are anonymised by rounding to the
nearest second. So these entries are close together, but not
necessarily at the same time.
How many DNS servers do you have configured?
(It looks like it's only one. That's quite a fragile
configuration.)
If it fails a request by chance, but the next request
succeeds, this is the pattern of messages you'll see.
Try adding a local caching resolver as the first listed
name server.
You might want to add your VPS DNS servers, and Google's
other server to the end of the list, too.
(A benefit of using local DNS servers is that fewer
networks see your DNS requests.
A drawback is that your VPS company then sees your DNS
requests and your traffic, but they could do this anyway.)