[Linux-ha-dev] Bind/stat Auth Patch

[Alan Robertson]

Matthew Berg wrote:
> On Tue, 2004-10-05 at 18:30, Matthew Berg wrote:
> 
> 
>>Also, it seems that we're not unlinking the path (at all? reliably?),
>>resulting in something like this:
>>
>>  [mberg at ashley heartbeat]$ ls *-*
>>  138f73b9-5cb6-c0f5-e37e-9306e007e9a4 
>>  8b7b5db0-0f49-c123-f7ce-afdd90a5a899 
>>  ea3e2180-4024-6db4-c0fd-d562f6e3dfde
>>  239dd81d-be62-c3af-c817-e78970b1cbc0 
>>  a95e2e20-9535-4531-bfae-d73a59fa24b8 
>>  eb9ad894-9e05-6b8c-badc-a7f016adeb3a
>>  2a28ccca-cffd-69cc-e03e-b7dcb2e3d468 
>>  ac17b796-e8d6-c88d-e866-d37037cbb9f6 
>>  eba98ab3-bb61-c2fe-d9dd-99d3d2e6cde9
>>  2e54149f-4694-ec22-aec5-f785abe2760a 
>>
>>I'll take a look at that tommorow as well.
> 
> 
> Just to reiterate what we talked about earlier (and for the benefit of
> people not on IRC), adding the unlink to socket_destroy_channel worked
> for this.

Hmmm... What socket is that?  What directories are these in?  That's really 
odd... And, the socket should be removed in socket_destroy_wait_conn() 
which is on the server side.  And, it *is* removed in that code.

I really don't understand what you're talking about here.  I don't think my 
IRC log is long enough to dig this out.



-- 
     Alan Robertson <alanr at unix.sh>

"Openness is the foundation and preservative of friendship...  Let me claim 
from you at all times your undisguised opinions." - William Wilberforce