[stunnel-users] stunnel client connecting to host with dynamic

Fri Feb 13 14:57:26 CET 2009

Sorry for messing things up.  I forgot to restart stunnel after adding 
relay = yes to the conf file.  It works OK for me now.

Regards,
Adamson H

-------- Original Message  --------
Subject: Re: [stunnel-users] stunnel client connecting to host with dynamic
From: C.J. Adams-Collier <cjac at colliertech.org>
To: Adamson H <adamson at polycastle.3322.org>
Date: 02/11/2009 02:59 AM
> If you watch port 53 (tcpdump, tshark, etc.), do you see a request
> leave the host and come back with the old name?  It sounds to me like
> you're getting a stale value from the cache somewhere...
>
> Can you run stunnel under strace or gdb and see what it does to get
> the hostname resolved?
>
> On Tue, Feb 10, 2009 at 11:08:36PM +0800, Adamson H wrote:
>   
>> Hi, I have named local resolver enabled on the stunnel client (FreeBSD  
>> 7.0). I did a test on the client and found there was no problem with the  
>> local resolver.
>>
>> 1. on the stunnel client: dig hostname.dynaip.org --> points to the  
>> correct IP of the stunnel host
>> 2. restart ppp on stunnel host to get a new IP
>> 3. on the stunnel client: dig hostname.dynaip.org --> points to the  
>> updated IP of the stunnel host
>>
>> I tested ftp service on the stunnel host without going through stunnel,  
>> and it worked before and after the IP change.
>>
>> Regards,
>> Adamson H
>>
>> -------- Original Message --------
>> Subject: Re: [stunnel-users] stunnel client connecting to host with dynamic
>> From: Michal Trojnara <Michal.Trojnara at mobi-com.net>
>> To: stunnel-users at mirt.net
>> Date: 02/10/2009 10:32 PM
>>     
>>> Adamson H wrote:
>>>   
>>>       
>>>> I added delay = yes to my stunnel client conf file and removed the 
>>>> cron script.  The rsync job (873) failed after the IP of the stunnel 
>>>> host had changed.  I had to manually restart stunnel to get the job 
>>>> done and put the cron script back again.
>>>>
>>>> delay = yes does not with work with ddns.
>>>>     
>>>>         
>>> That's really strange.  The relevant client.c code is:
>>>
>>>     /* setup address_list */
>>>     if(c->opt->option.delayed_lookup) {
>>>         resolved_list.num=0;
>>>         if(!name2addrlist(&resolved_list,
>>>                 c->opt->remote_address, DEFAULT_LOOPBACK)){
>>>             s_log(LOG_ERR, "No host resolved");
>>>             longjmp(c->err, 1);
>>>         }
>>>         address_list=&resolved_list;
>>>     } else /* use pre-resolved addresses */
>>>         address_list=&c->opt->remote_addr;
>>>
>>> Maybe you use some kind of advanced/broken resolver library that caches IP
>>> addresses?
>>>
>>> Best regards,
>>>     Mike
>>> _______________________________________________
>>> stunnel-users mailing list
>>> stunnel-users at mirt.net
>>> http://stunnel.mirt.net/mailman/listinfo/stunnel-users
>>>
>>>   
>>>       
>> _______________________________________________
>> stunnel-users mailing list
>> stunnel-users at mirt.net
>> http://stunnel.mirt.net/mailman/listinfo/stunnel-users
>>     