| Title: | DEC X.25 for OpenVMS AXP |
| Moderator: | OZROCK::MUGGERIDGE |
| Created: | Mon Jan 18 1993 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 524 |
| Total number of notes: | 2218 |
X25 1.0g DECnet OSI V6.3 and OpenVMS Alpha V6.2-1H3
---------------------------------------------------
Customer has VAX multihost with several VAX access nodes working OK. The only
Alpha connected to this have BIG problems when trying to use X29.
One or two users can normally connect to it but then calls will be disconnected
after PW is typed in(or in Username if you keep on typing 50-60 characters.
Example:
SNORRE> set h/x29 <customer_dte_#>
%PAD-I-COM, call connected to remote DTE
Username: FIELD
Password:
Error reading command input
Timeout period expired
%PAD-I-CLEAR, call being cleared at request of remote DTE
PAD usage statistics:
Elapsed time: 00:00:25.88 CPU time: 00:00:00.20
Buffered I/O: 102 Direct I/O: 70
Page faults: 75
YES, I HAVE typed <Return> after the PW.
The customer claims he could get 6+ successful logins on an idle system, but we
never succeed more than one or two during normal working hours so it seems like
it could be a resource problem of some kind. The Alphaserver is a 4100 with 2GB
of memory with about 700 processes during normal working hours.
I've traced the login on my local system and on the Multihost VAX system at
customer site and all characters seems to be send and received OK(including the
<Return>). BUT I'm unable to trace it on the Alpha because NETTRACE.EXE is not
there and when renaming the existing NETTRACE_TV.EXE it complains about some old
shareable images in SYS$LIBRARY. It seems like it's a mess of new and old images
related to NETTRACE on the Alpha.
Could anyone tell me how I could use TRACE on the Alpha?
Any hints on how to troubbleshoot the login problem?
\�ge
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 515.1 | DEcnet PLUS trace utility is CTF | OZROCK::HARTWIG | Arthur Hartwig, TaN Engineering-Australia | Sat May 03 1997 08:16 | 15 |
In the DECnet/OSI environment CTF is the trace utility. CTF is invoked
by the command TRACE.
This is all written up fairly clearly in the DECnet/OSI documentation.
The command TRACE START/LIVE "X25GAP CIRCUIT *" will possibly provide
some useful information. You might also look for network errors causing
timeouts and re-transmissions. Do all the systems have sufficient
non-paged pool?
I have successflly used a VAX/VMS system running DECnet/OSI as a X.25
gateway for Alpha VMS systems. So there is some sort of precedent for
what you are trying to do.
| |||||
| 515.2 | NO nettrace.exe on this system... | OSLAGE::AGE_P | Aage Ronning, Oslo, Norway, (DTN 872-8464) | Mon May 05 1997 19:34 | 12 |
>> In the DECnet/OSI environment CTF is the trace utility. CTF is invoked >> by the command TRACE. The problem is that TRACE tries to invoke sys$system:nettrace.exe, BUT on this Alpha system there is NO nettrace.exe(only a nettrace_tv.exe). I tried to rename this image, but then it seemed like the sys$library files where incompatible with the image. I don't have access to the system right now, but I could post the output later if wanted. What is the easiest way to get trace up and working (Could I copy some files)? \�ge | |||||
| 515.3 | TRACE shouldn't try to invoke NETTRACE! | OZROCK::HARTWIG | Arthur Hartwig, TaN Engineering-Australia | Mon May 05 1997 22:09 | 7 |
Are you sure the system has X.25 V1.0-G installed. It sounds as if X25
CLIENT (a different product) is at least partially installed. What does
$ product show product
say?
| |||||
| 515.4 | 7 NW devices owned by "X25$APPLD" ? | OSLAGE::AGE_P | Aage Ronning, Oslo, Norway, (DTN 872-8464) | Tue May 06 1997 18:03 | 97 |
$ prod show prod
PRODUCT
-------
DEC AXPVMS DECNET_OSI V6.3
DEC AXPVMS DWMOTIF V1.2-3
DEC AXPVMS NASAXP V8.0
DEC AXPVMS OSAK V3.0-L
DEC AXPVMS UCX V4.1-12
DEC AXPVMS VMS V6.2
DEC AXPVMS WANDD V1.0-G
DEC AXPVMS X25 V1.0-G
%PCSI-I-IOPPDB, additional information last updated 16-MAR-1997 in DISK$AXPVMSSY
S:
$ sh mem/pool/full
System Memory Resources on 6-MAY-1997 08:52:14.84
Nonpaged Dynamic Memory (Lists + Variable)
Current Size (bytes) 24993792 Current Size (pagelets) 48816
Initial Size (NPAGEDYN) 24993792 Initial Size (pagelets) 48816
Maximum Size (NPAGEVIR) 99999744 Maximum Size (pagelets) 195312
Free Space (bytes) 6432576 Space in Use (bytes) 18561216
Size of Largest Block 1689728 Size of Smallest Block 64
Number of Free Blocks 16615 Free Blocks LEQU 64 Bytes 2680
Free Blocks on Lookasides 7613 Lookaside Space (bytes) 1716096
Paged Dynamic Memory
Current Size (PAGEDYN) 15884288 Current Size (pagelets) 31024
Free Space (bytes) 1664096 Space in Use (bytes) 14220192
Size of Largest Block 838448 Size of Smallest Block 16
Number of Free Blocks 25338 Free Blocks LEQU 64 Bytes 25336
$
Why do we have 7 NW devices owned by "X25$APPLD" even seconds after a restart?
$ sh dev/full nw
Device NWA0:, device type X.25 datalink, is online, mailbox device, device is a
template only.
Error count 0 Operations completed 1
Owner process "" Owner UIC [SYSTEM]
Owner process ID 00000000 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 0 Default buffer size 0
Device NWA79:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 0
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA80:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA81:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA82:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA83:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA84:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
Device NWA85:, device type X.25 datalink, is online, mailbox device.
Error count 0 Operations completed 1
Owner process "X25$APPLD" Owner UIC [DATA,ARNFINN_WA]
Owner process ID 21033087 Dev Prot S:RWPL,O:RWPL,G,W
Reference count 1 Default buffer size 0
$
\�ge
| |||||
| 515.5 | MARVIN::CARLINI | Tue May 06 1997 21:31 | 13 | ||
>The problem is that TRACE tries to invoke sys$system:nettrace.exe, BUT on this >Alpha system there is NO nettrace.exe(only a nettrace_tv.exe). I tried to >rename As Arthur notes, NETTRACE is *not* what you should expect TRACE to invoke! I have hit this problem once, recently on a VAX. Something, somewhere had installed the old PSI Phase IV TRACE utility (NETTRACE); in my case it was probably an installation of an X25router kit or somesuch. How this might happen on an Alpha, I have no idea. Try a re-install of DECnet-Plus followed by a reboot, if that does not sort out TRACE, send me mail and I'll send you the TRACE.CLD source as a workaround. Antonio | |||||
| 515.6 | OZROCK::HARTWIG | Arthur Hartwig, TaN Engineering-Australia | Thu May 08 1997 05:04 | 9 | |
There is an NW device owned by X25$APPLD for EACH X.25 or X.29
application that has been defined (by e.g. $ mcr ncl create x25 access
application ... ).
X25$APPLD does NOT own NW devices for x25 applications that are not
created dynamically in response to incoming calls. (X29 login is the
single exception.)
| |||||
| 515.7 | NSP flow control policy again? | UTRTSC::VROOIJ | Learning to fly... | Tue May 13 1997 20:36 | 10 |
For the logins problems have you checked 'NSP FLOW CONTROL POLICY'
The last few weeks I have had several customers complaining about
X25GAP problems from alpha with X25gateway/DECnis as connector node.
As soon as the flow control policy is changed from segment to NO flow
control the problems where gone. I haven't had time to build a test
environment to reproduce the problem but I believe on Alpha there is
still a flow control problem with 'segment flowe control'
Henk
| |||||
| 515.8 | OZROCK::HARTWIG | Arthur Hartwig, TaN Engineering-Australia | Wed May 14 1997 00:10 | 3 | |
Have you lodged an IPMT on the flow control issue?
| |||||
| 515.9 | No IPMT yet. | UTRTSC::VROOIJ | Learning to fly... | Thu May 15 1997 21:16 | 8 |
Arthur,
I haven't logged an IPMT yet. All I know at this moment is that
changing the flow control solves the customer problem. If time permits
I want to setup an test-environment to see if it can be reproduced.
I'll log an IPMT then.
Henk
| |||||