Date: Tue, 20 Jul 93 04:30:06 PDT
From: Advanced Amateur Radio Networking Group <tcp-group@ucsd.edu>
Errors-To: TCP-Group-Errors@UCSD.Edu
Reply-To: TCP-Group@UCSD.Edu
Precedence: Bulk
Subject: TCP-Group Digest V93 #185
To: tcp-group-digest
TCP-Group Digest Tue, 20 Jul 93 Volume 93 : Issue 185
Today's Topics:
JNOS 1.10x5 LZW problem
Send Replies or notes for publication to: <TCP-Group@UCSD.Edu>.
Subscription requests to <TCP-Group-REQUEST@UCSD.Edu>.
Problems you can't solve otherwise to brian@ucsd.edu.
Archives of past issues of the TCP-Group Digest are available
(by FTP only) from UCSD.Edu in directory "mailarchives".
We trust that readers are intelligent enough to realize that all text
herein consists of personal comments and does not represent the official
policies or positions of any party. Your mileage may vary. So there.
----------------------------------------------------------------------
Date: Mon, 19 Jul 93 19:21:04 BST
From: john@its.bt.co.uk
Subject: JNOS 1.10x5 LZW problem
To: tcp-group@ucsd.edu
Hi,
Can anybody answer this question. Either mail the reply to me or sent
it to tcp-group and I can forward the reply.
John
--- Forwarded mail from g7lob@gb7lob.#11.gbr.eu
Date: 18 Jul 93 14:41:00 UTC
From: g7lob@gb7lob.#11.gbr.eu
To: tcpip@gbr
Subject: JNOS 1.10x5 .. LZW problem.
R:930718/1441Z @:GB7LOB.#11.GBR.EU [TCP/IP Gateway] #:4681 $:4681_GB7LOB
From: G7LOB@GB7LOB.#11.GBR.EU
To : TCPIP@GBR
Hi all,
I am trying out JNOS 1.10x5 here and have (I think) established the cause of
what was for me a major problem.
I run JNOS in a DesqView window, along with FBB, and they forward to/from
each other internally via a BPQ switch. Despite having loaded the whole world
into high memory and compiled JNOS with only the bits I find I really need, on
startup I have around 100k of free memory (coreleft as indicated by the latest
JNOS on the prompt line); this free memory falls rapidly as NOS establishes
itself.. then stays stable at around 40-50k ((or it does now!!)).
With smtp LZW compression enabled, I was finding thatwhen my FBB-system
received a bundle of bulletins/messages from the NTS all together, they were
forwarded to JNOS (BBS-forwarding) where the were then put into the smtp queue
for rewrite/alias/onward smtp transmission. Seemingly the LZW compression
routines were swallowing up every last bit (byte?) of free memory, such that if
I did an 'smtp kick', the repeatedly hit the return key, I could watch the free
core memory shrink down.. 20k..10k..2k..512bytes.. 10 bytes (honest!) **CRASH**
Obviously the 'memory threshold' setting (16384 on here) was unable to stop
the smtp (or LZW compression) process from grabbing this memory.
I have now turned LZW compression off ("smtp sendlzw off : smtp reclzw off")
and the problem seems to have disappeared.. in addition the processing of the
messages seems much faster (maybe that's obvious!).. The only slight
disadvantage now is that eavesdroppers can read all our smtp mail!.
I am a (so-called!) 'C' programmer, but seldom get time to delve deeply
enough into the NOS sources to even attempt to debug this one! .. Maybe this
problem has already been identified (and a fix posted??) on the Internet??
Maybe someone with internet access can extract the relevant bits from this
message and post it??
George, G7LOB, sysop @ GB7LOB.#11.GBR.EU
TCP/IP BBS GB7LOB-5 on 144.625MHz [44.131.3.110]
--- End of forwarded message from g7lob@gb7lob.#11.gbr.eu
--
+------------------------------------+------------------------------------+
+ Work + Play +
+ Internet: jvt@its.bt.co.uk + Internet: john@its.bt.co.uk +
+ + Amprnet: john@g4rev.ampr.org +
+ + BBS: G4REV@GB7LOB.#11.GBR.EU +
+------------------------------------+------------------------------------+
+ Intel free zone +
+-------------------------------------------------------------------------+
------------------------------
End of TCP-Group Digest V93 #185
******************************
******************************