[nos-bbs] Bleeding add-on errors

[Michael Fox - N6MEF]

Thanks Bob.  Nice catch!

This may be related:

I don't have bleeding(1 or2) on production machines.  But I have bleeding on one machine and bleeding2 on another machine, both of which talk to production machines.  Not long after upgrading one of them to bleeding2, two of our five production machines had a weird problem.  The area that holds ww bulletins grew to an enormous size (6+ GB, when it is usually 1-2 MB).  When indexing occurred, the area.new file was about the same size, filling up the partition.  I have JNOS is in its own partition, specifically to mitigate issues like this.  So the rest of the machine continue to function and I could get in and see what was happening.

Looking at the messages in the bulletin ww area, many had the same content repeated over and over many, many times, making some messages very, very long.  Attempts to delete problem messages caused JNOS to crash.  And, being that these were production machines, I elected to just delete the .txt, .new and .idx files for that area in order to restore operations quickly.

I've been monitoring closely since then and the problem hasn't returned.  But the timing was curious, since I haven't seen anything like that in 6+ years.

Just speculation.  But perhaps the corrupted From line caused the index to get confused, leading to other problems...

I guess I should roll back those two machines to 2.0j.7v until we hear more from Maiko about what's next.

Michael
N6MEF



> -----Original Message-----
> From: nos-bbs [mailto:nos-bbs-bounces at tapr.org] On Behalf Of Boudewijn
> (Bob) Tenty
> Sent: Tuesday, April 26, 2016 9:19 PM
> To: TAPR xNOS Mailing List <nos-bbs at tapr.org>
> Subject: [nos-bbs] Bleeding add-on errors
> 
> Forwarding.
> 
> Effected: From/From: addresses in the headers and missing R line.
> 
> Be aware that the bleeding stuff can screw up the From: and From addresses
> in the headers and it doesn't add the R line of the last forwarding
> station.
> That happens when one of the stations is running a bleeding version and
> the
> other jnos without the bleeding stuff. (or when both are running it)
> 
> It doesn't happen (As far as I can see, so no guarantees) when you
> are running the bleeding version and the far end is running tnos 2.30,
> fbb or bpq.
> 
> I earlier informed Maiko about it.
> 
> Example. (ve3mch to va3bal but it also happens the other way around)
> 
> #scon.on.can.noam is not appended to ve3mch (before  the '@' in line 1)
> From: Completely messed up, should be ve3tok at ve3mch.#scon.can.noam
> R line is missing
> 
> 
>  From ve3tok%ve3mch at va3bal.ampr.org Tue Apr 26 21:13:44 2016
> Received: from va3bal.ampr.org by va3bal.ampr.org (JNOS2.0j.7vBL) with
> SMTP
>      id AA836 ; Tue, 26 Apr 2016 21:13:44 EDT
> Date: Tue, 26 Apr 2016 21:13:41 EDT
> Message-Id: <1359_ve3mch at ve3mch.bbs>
> From: ve3tok%ve3mch.bbs at va3bal.ampr.org
> To: ve3tok at va3bal.#scon.on.can.noam
> Subject: g
> X-BBS-Msg-Type: E
> X-JNOS-User-Port: Circuit  (HAMS:VE3MCH-6 VE3MCH-7)  -> Sending message
> Status: R
> 
> bleed
> 
> 
> Don't use the bleeding stuff in a production environment.
> 
> Bob VE3TOK
> 
> 
> _______________________________________________
> nos-bbs mailing list
> nos-bbs at tapr.org
> http://www.tapr.org/mailman/listinfo/nos-bbs