[tor-relays] stem library: huge difference between resolvers "proc" and others, eg. "netstat"

[Toralf Förster]

On 03/10/2018 01:05 AM, Damian Johnson wrote:
> Unfortunately I can't troubleshoot this without a local repro. If
> you'd care to dig in I'd suggest adjusting the script a little to
> print the connections, then see in what way netstat differs from proc.
> Is it a strict superset? Does it have duplicates? Anything about the
> connections which differ that seem interesting?
> 
> Cheers! -Damian

Well, comparing the results of "proc" and "netstat" via
	print ("debug: %s:%i\t%s:%i" % (laddr, lport, raddr, rport) ))

shows these things :

1. both have few duplicate lines (2-3 for netstat, about 6 for proc)
2. proc expands the ipv6 addresses, eg.: 2a01:4f8:190:514a::2 into 2a01:04f8:0190:514a:0000:0000:0000:0002
3. the resolver netstat have significant lines more w/o any patterns why
4. proc is the only resolver which differs, lsof, netstat and ss have comparable results
5. the output of netstat is nearly identical to "na -n | wc -l"

-- 
Toralf
PGP C4EACDDE 0076E94E

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 228 bytes
Desc: OpenPGP digital signature
URL: <http://lists.torproject.org/pipermail/tor-relays/attachments/20180317/72a6835e/attachment.sig>