http://www.slony.info/bugzilla/show_bug.cgi?id=55

           Summary: Possible buffer overflow in slon, still happens as of
                    1.2.14
           Product: Slony-I
           Version: 1.2
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: minor
          Priority: low
         Component: slon
        AssignedTo: slony1-bugs at lists.slony.info
        ReportedBy: ahodgson at simkin.ca
                CC: slony1-bugs at lists.slony.info
   Estimated Hours: 0.0


Once in a while, I experience a problem replicating a new set, where the slon
remote worker thread is terminated by a signal 11.

I'm not much of a C programmer, so I haven't been able to patch this myself.

I don't know how much of the following is relevant, but here is what I do know:

It only occurs when debugging is set at 2 or higher. 

According to strace, the last thing the process does before receiving the sig
11 is output the log entry "syncing set %d with %d table(s) from provider %d\n"
(with real data of course), which appears to be called from remote_worker.c at
line 4232. The log entry is fully output before the thread is killed.

In my case, it always happens on set id 100, which I use as a temporary set
when adding new tables and sequences, and only on the first sync of the set
after subscribing it to a new node.

Reducing the debugging to 0 has so far gotten past the issue all 3 times it has
happened, and later syncs at a higher debug level work fine. It's not
reproducible except at the specific time it's happening (ie. I don't have a
test case that can demonstrate it).

I've looked at the logging code and can't see a buffer issue in it. I think
probably it's just causing an interaction with the real problem elsewhere.


-- 
Configure bugmail: http://www.slony.info/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
You are the assignee for the bug.