#20423: Clock jumps on relay due to hostname lookup timeouts
--------------------------+------------------------------------
Reporter: Felixix | Owner:
Type: defect | Status: needs_review
Priority: High | Milestone: Tor: 0.2.9.x-final
Component: Core Tor/Tor | Version: Tor: 0.2.8.9
Severity: Normal | Resolution:
Keywords: regression | Actual Points:
Parent ID: | Points:
Reviewer: | Sponsor:
--------------------------+------------------------------------
Changes (by arma):
* keywords: regression freebsd => regression
* status: new => needs_review
Comment:
My {{{bug20423}}} branch should resolve this issue.
(It does not resolve the fact that maybe the relay has a broken resolver,
but it will at least not try to do thousands of resolve attempts in a
row.)
It is a patch on maint-0.2.9. We can consider a backport to 028 if we like
it enough, or we can leave 028 as a regression and focus on our future.
Once we get this one in, we might in 0.3.0 consider changing many more of
these calls to cache_only=1. In fact, if we leave
check_descriptor_ipaddress_changed() calling resolve_my_address() once a
minute, maybe we're all set and the whole of
router_pick_published_address() can just look at the cached values? I
didn't want to make an invasive change like that in 0.2.9 though, since
there are probably edge cases we care about, e.g. where
getinfo_helper_misc() calls router_pick_published_address() and we're not
a relay so we've never called resolve_my_address() before that very
moment.
--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/20423#comment:31>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online
_______________________________________________
tor-bugs mailing list
tor-bugs@lists.torproject.org
https://lists.torproject.org/cgi-bin/mailman/listinfo/tor-bugs
