BIND 9 - I'm not sure is working correctly!

Mark netx at softhome.net
Sun May 20 18:21:53 UTC 2001



> Hmmm.  Interesting.
> In the second case, if you do a simple:
>
> ## Rh 7.1 (with BIND 9.1.1)
> # ping -c 25  193.230.52.100
>
> Do you get any packet loss?

NO (because in this case BIND is not involved, I think)

> I'm really kinda confused about the "-n" option.  I'm not sure what:
>    "No attempt will be made to lookup
>     symbolic names for host addresses.
>    "
> really means -- on my RH6.2 system I get the same output with or without
> "-n" :|

You are absolutely right ! As I said in RH 7.0 or previous I also get the
same
result with or without adding "-n" at ping.

Now, Why I said I'm not sure ? :

I've installed a clean RH 7.1 (BIND 9.1.0), then I added my config and zone
files (which I verified and they are correct) then I observed that packet
lose at
"ping www.mydomain.com" then I upgraded to BIND 9.1.1, .... the same
problem. Then I added "in-addr.arpa zone" (just to test because normally
I can not do that because I don't have an entire C class) and then I issue
"ping -c 25 www.mydomain.com" and that's working just fine. So, If I remove
"in-addr.arpa zone" is working fine, if and only if I issue "ping -n".

So you may think this is in fact regarding to ping. Maybe, maybe not, Why ?
I downgraded to BIND 8.2.3 (on RH 7.1) and guess what.... there is no packet
lose when I issue a simple "ping www.mydomain.com", ("-n" option is not
necessary) !

Then I try this: I installed a clean RH 7.0 and upgrade only BIND to 9.1.1
and there is no packet lost (there is no difference between "ping -n
www.mydomain.com" and ping www.mydomain.com. Hmm.

I think this problem is confusing enough, I've checked docs, FAQ, and this
list archive and no one reported this problem. Now I know why:

If BIND 9 runs on RH previous to 7.1 there is no problem. They are going to
see this problem when they will start running BIND 9 on RH 7.1.

P.S. BIND 9 on Mandrake 8 have the same problem.

So, any idea ???

>
> I'm asking this because I just upgraded from RH 7.0 to RH 7.1. BIND
9.1.0/1
> is
> acting very strange on RH 7.1 (having exactly the same config and zone
> files).
>
> On RH 7.0 (with BIND 9.1.1)     ->           ping -c 3 www.comefnet.com
>
> PING www.comefnet.com (193.230.52.100) from 193.230.52.100 : 56(84) bytes
of
> data.
> 64 bytes from 193.230.52.100: icmp_seq=0 ttl=255 time=315 usec
> 64 bytes from 193.230.52.100: icmp_seq=1 ttl=255 time=90 usec
> 64 bytes from 193.230.52.100: icmp_seq=2 ttl=255 time=79 usec
>
> --- 193.230.52.100 ping statistics ---
> 3 packets transmitted, 3 packets received, 0% packet loss
> round-trip min/avg/max/mdev = 0.079/0.161/0.315/0.109 ms
>
>
> On Rh 7.1 (with BIND 9.1.1)       ->        ping -c 25 www.comefnet.com
>
> PING www.comefnet.com (193.230.52.100) from 193.230.52.100 : 56(84) bytes
of
> data.
> 64 bytes from 193.230.52.100: icmp_seq=0 ttl=255 time=358 usec
> 64 bytes from 193.230.52.100: icmp_seq=1 ttl=255 time=9.030 sec
> 64 bytes from 193.230.52.100: icmp_seq=2 ttl=255 time=18.060 sec
> 64 bytes from 193.230.52.100: icmp_seq=3 ttl=255 time=23.712 sec
> 64 bytes from 193.230.52.100: icmp_seq=4 ttl=255 time=32.730 sec
> 64 bytes from 193.230.52.100: icmp_seq=5 ttl=255 time=41.750 sec
>
> --- www.comefnet.com ping statistics ---
> 25 packets transmitted, 7 packets received, 72% packet loss <-----------
> ?????????
> round-trip min/avg/max/mdev = 0.358/25150.562/50770.145/16622.264 ms
>
> I think the ping binary is also modified because is working fine IF AND
ONLY
> IF:
>
> 1. I add "in-addr.arpa" zone for my domain (very hard because is not an
> entire C class) or
> 2. I use ping -n www.mydomain.com (-n disable host lookup) or
> 3. I use BIND 9.1.1 on RH 7.0 and use the ping that comes with RH 7.0.
>
> So "ping" that comes with RH 7.1 needs "-n" to work.
>
> "nslookup" and "dig" are working fine returning correct info, and LAN
> clients don't have
> problems with DNS and they are UNIX and Win9x.
>
> After all this my question is:
>
> How can I be sure BIND 9 is working correctly ?
>
>
> Mark
>
>
>
>





More information about the bind-users mailing list