Weird Bind issue

Travis Roy travis at scootz.net
Tue Feb 8 13:41:01 EST 2005


Ahh, think you very much Jeff.

We are running the latest rpm package of bind, but it is one that 
suffers from these bugs.

We'll probably have to just roll our own.

> On Tue, Feb 08, 2005 at 12:31:03PM -0500, Travis Roy wrote:
> 
>>This started cropping up last week. Our DNS server is running fine, then 
>>for no reason it dies with this error:
>>
>>exiting (due to assertion failure)
> 
> 
> What version of Bind ? 
> These bugs were fixed in 9.2.0a1 release:
> 
>  720.	[bug]	Server could enter infinite loop in
> 		dispatch.c:do_cancel(). [RT #733]
> 
>  719.	[bug]	Rapid reloads could trigger an assertion failure.
> 		[RT #743, #763]
> 
> http://linux-howto.net/usr/doc/bind-9.2.2-P3/CHANGES
> 
> 
> More info: [From Trav's offlist post to me in response to my
> mis-addressed private post (had intended to post to the list)]
> 
> Feb  6 06:56:51 auth00 named[9254]: adb.c:2755: REQUIRE(findp != ((void
> *)0) && (((*findp) != ((void *)0)) && (((const isc__magic_t
> *)(*findp))->magic == ((('a') << 24 | ('d') << 16 | ('b') << 8 |
> ('H')))))) failed
> Feb  6 06:56:51 auth00 named[9254]: exiting (due to assertion failure)
> 
> Feb  8 06:09:29 auth00 named[15541]: dispatch.c:1706: REQUIRE(dispp !=
> ((void *)0) && ((((*dispp)) != ((void *)0)) && (((const isc__magic_t
> *)((*dispp)))->magic == ((('D') << 24 | ('i') << 16 | ('s') << 8 |
> ('p')))))) failed
> Feb  8 06:09:29 auth00 named[15541]: exiting (due to assertion failure)
> 
> Feb  8 11:06:33 auth00 named[15689]: rbtdb.c:2819: REQUIRE(((rbtdb) !=
> ((void *)0) && (rbtdb)->common.impmagic == (('R') << 24 | ('B') << 16 |
> ('D') << 8 | ('4')))) failed
> Feb  8 11:06:33 auth00 named[15689]: exiting (due to assertion failure)
> 
> 
> All three of these asserts check for a null pointer as the first thing.
> 
> But based on the line numbers I would say you may be on some version of
> Bind later or earlier than 9.2.4.   Sorry - not very helpful.
> 
> 
> 
> 
>>Anybody have any ideas, we've been plugging away at this problem for 3 
>>days now, it's getting VERY annoying.
>>_______________________________________________
>>gnhlug-discuss mailing list
>>gnhlug-discuss at mail.gnhlug.org
>>http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
>>
> 
> 




More information about the gnhlug-discuss mailing list