On Wed, 15 May 2013, G. wrote:
On Wed, 15 May 2013 22:46:10 -0000, you wrote:
?Stopcode KNIKSP, type=INFO, on CPU0 at 15-May-113 18:43:58
[Continuing system]
Translation:
Name Module Type Message and Explanation
------ ------ ---- ---------------------------------------------
KNIKSP KNISER INFO KLNI Service Paused. The maximumm number of
packets that can be processed has been
exceeded. Location .PBMPP (Maximum Packets
Processed) in the KDB/PCB contins the limit,
and that limit is 2000 (decimal).
It's idling...but by idling it's stopping completely. ;)
Strange. I've seen that before when I had problems with Ethernet. Anyway,
being an INFO stopcode, it should be just, well, informational and the system
should continue as usual. Or is it throwing that stopcode again and again?
It's throwing it again and again. Every single time Ethernet drops. ;)
My Ethernet is configured like this:
devdef ni0 564 ni20 dedic=0 decnet=1 doarp=0 enaddr=AA:00:04:00:xx:xx dpdelay=12 c3dly=2 rdtmo=3
Yours?
devdef ni0 564 ni20 dedic=1 decnet=1 doarp=0 enaddr=aa:00:04:00:aa:24 ifc=eth0 dpdelay=12 c3dly=10 rdtmo=3
I'll try changing c3dly to 2.
G.
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
On Wed, 15 May 2013 22:46:10 -0000, you wrote:
?Stopcode KNIKSP, type=INFO, on CPU0 at 15-May-113 18:43:58
[Continuing system]
Translation:
Name Module Type Message and Explanation
------ ------ ---- ---------------------------------------------
KNIKSP KNISER INFO KLNI Service Paused. The maximumm number of
packets that can be processed has been
exceeded. Location .PBMPP (Maximum Packets
Processed) in the KDB/PCB contins the limit,
and that limit is 2000 (decimal).
It's idling...but by idling it's stopping completely. ;)
Strange. I've seen that before when I had problems with Ethernet. Anyway,
being an INFO stopcode, it should be just, well, informational and the system
should continue as usual. Or is it throwing that stopcode again and again?
My Ethernet is configured like this:
devdef ni0 564 ni20 dedic=0 decnet=1 doarp=0 enaddr=AA:00:04:00:xx:xx dpdelay=12 c3dly=2 rdtmo=3
Yours?
G.
On Wed, 15 May 2013, G. wrote:
On Wed, 15 May 2013 21:49:14 -0000, you wrote:
Oh, and there is the patch to enable the idle feature of KLH10 too!
That will be even more helpful!
Patching monitor sources, rebuilding monitor modules and then replacing them
into TOPS10.REL is a bit involved and now here is late evening, so remember me
tomorrow or in a couple of days. :)
For now I'll just write how to patch the monitor file because it's quickier.
First of all, add the following to your KLH10 initialization file:
devdef idler 740 host
Then make a copy of your SYSTEM.EXE and work on the copy:
..R FILDDT
File: xxxxxx.EXE/P <--- /p to enable patching
[Patching file DSKB:xxxxxx.EXE[1,2]]
[35118 symbols loaded from file]
[ACs copied from CRSHAC to 0-17]
<--- FILDDT prints nothing, but it's ready to accept commands
nulcod+1/ <--- type this and nothing more
nulcod+1/ SOJG W,P <--- DDT will show you the contents of that location
nulcod+1/ SOJG W,P CONO 740,1 <--- tell him to put this there instead and
press enter to confirm
<--- if all goes well FILDDT prints nothing again (friendly, isn't it?)
^Z <--- to save and exit
Now try to boot with your patched monitor and see if does idle KLH10. This
monitor will work only if there is a device at address 740.
That didn't quite work as planned...
?Stopcode KNIKSP, type=INFO, on CPU0 at 15-May-113 18:43:58
[Continuing system]
It's idling...but by idling it's stopping completely. ;)
G.
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
On Wed, 15 May 2013 21:47:12 -0000, you wrote:
RUN BATCON seems to not start. It might need started earlier though.
BATCON cannot be RUNned. It gets started by Galaxy when needed.
Try this:
.copy test.ctl=tty:
.dir <--- you have to type the dot too!
^Z
.submit/notify/out:nolog test
[Batch job TEST queued, request #58, limit 0:05:00]
.
[From SYSTEM: Job TEST request #58 finished executing at 0:32:40]
If everything goes well, you'll receive the completion message and there will
be a TEST.LOG file in your directory. TYPE it to see if it's all OK.
/notify let's you know when the job finishes, /output:nolog prevents queueing
the log for printing (given that there is no printer emulation on KLH10...).
HTH,
G.
On Wed, 15 May 2013 21:49:14 -0000, you wrote:
Oh, and there is the patch to enable the idle feature of KLH10 too!
That will be even more helpful!
Patching monitor sources, rebuilding monitor modules and then replacing them
into TOPS10.REL is a bit involved and now here is late evening, so remember me
tomorrow or in a couple of days. :)
For now I'll just write how to patch the monitor file because it's quickier.
First of all, add the following to your KLH10 initialization file:
devdef idler 740 host
Then make a copy of your SYSTEM.EXE and work on the copy:
.R FILDDT
File: xxxxxx.EXE/P <--- /p to enable patching
[Patching file DSKB:xxxxxx.EXE[1,2]]
[35118 symbols loaded from file]
[ACs copied from CRSHAC to 0-17]
<--- FILDDT prints nothing, but it's ready to accept commands
nulcod+1/ <--- type this and nothing more
nulcod+1/ SOJG W,P <--- DDT will show you the contents of that location
nulcod+1/ SOJG W,P CONO 740,1 <--- tell him to put this there instead and
press enter to confirm
<--- if all goes well FILDDT prints nothing again (friendly, isn't it?)
^Z <--- to save and exit
Now try to boot with your patched monitor and see if does idle KLH10. This
monitor will work only if there is a device at address 740.
G.
On Wed, 15 May 2013, Cory Smelosky wrote:
On Wed, 15 May 2013, G. wrote:
On Wed, 15 May 2013 20:40:01 -0000, you wrote:
Disregard, I made some changed to accounting and everything seems
operational now. I can't SET KSYS, but every single other thing works!
KSYS requires BATCON, so maybe there is something wrong with that. Do you have
tweaked your SYS:SYSTEM.CMD file and set the right permissions?
It should contain at least something like this to enable batch processing:
SET BATCH-STREAM 0 TIME-LIMIT 0:10
SET BATCH-STREAM 1 TIME-LIMIT 0:60
SET BATCH-STREAM 2 TIME-LIMIT 10:100000
START BATCH-STREAM 0:2
SET BATCH-STREAM 0:3 NOOPR-INTERVENTION
SET BATCH-STREAM 0 TIME-LIMITS 0:300
SET BATCH-STREAM 1 TIME-LIMITS 0:600
SET BATCH-STREAM 2 TIME-LIMITS 600:6000
SET BATCH-STREAM 3 TIME-LIMITS 0:6000
SET BATCH-STREAM 3 PRIORITY 20:63
START BATCH-STREAM 0:3
is the system.cmd i'm using (stolen from TWONKY)
RUN BATCON seems to not start. It might need started earlier though.
Disregard. I seem to just be impatient.
And something like this to enable remote DECnet file access:
SET FAL-STREAM 0 NETWORK DECNET
SET FAL-STREAM 1 NETWORK DECNET
SET FAL-STREAM 2 NETWORK DECNET
DEFINE FILE-ACCESS DEFAULT-PPN [377777,377777]
START FAL-STREAM 0:2
Those I WILL add though.
You may want to try these commands at the OPR> prompt before putting them into
the SYS:SYSTEM.CMD file.
Status of MARLEY TOPS-10 MONITOR at 16:39:09 on 15-May-113
Uptime 43:17, 114% Null time = 114% Idle + 0% Lost, 1% Overhead
More than 100% is not very good. It happened to me with a monitor that was
unable to keep time: try some DAYTIME commands in a couple of minutes and
check if it says the correct wall clock time or not...
It's off by about 30 seconds to a minute...not unexpected due to manually
setting the time.
9 [OPR] 751 5 ^C 0
This does not seem very good too: it's not detached, it's running nothing, and
it's in "^C" state... Maybe it's a stale BATCON process?
Yeah, that's a stale job.
HTH,
G.
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
On Wed, 15 May 2013, G. wrote:
On Wed, 15 May 2013 22:38:35 +0200, you wrote:
There is a section in SIG.MEM that explains very clearly and easily how to
create a complete system backup on tape.
To Cory:
I reply to myself because I remembered an important thing: the TOPS-10 V7.05
monitor has a bug that prevents it from initializing new structures when doing
a scratch install. So, if you create a recovery tape containing that monitor,
you'll not be able to initialize disks and proceed with the installation.
The solution I found is a simple patch that adds a single instruction in a
monitor module which got modified during the very last development cycle and
was probably never throughly tested. When you'll be ready and with some spare
time, remember me to tell you how and what to do! :)
That'll be helpful.
Oh, and there is the patch to enable the idle feature of KLH10 too!
That will be even more helpful!
G.
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
On Wed, 15 May 2013, G. wrote:
On Wed, 15 May 2013 20:40:01 -0000, you wrote:
Disregard, I made some changed to accounting and everything seems
operational now. I can't SET KSYS, but every single other thing works!
KSYS requires BATCON, so maybe there is something wrong with that. Do you have
tweaked your SYS:SYSTEM.CMD file and set the right permissions?
It should contain at least something like this to enable batch processing:
SET BATCH-STREAM 0 TIME-LIMIT 0:10
SET BATCH-STREAM 1 TIME-LIMIT 0:60
SET BATCH-STREAM 2 TIME-LIMIT 10:100000
START BATCH-STREAM 0:2
SET BATCH-STREAM 0:3 NOOPR-INTERVENTION
SET BATCH-STREAM 0 TIME-LIMITS 0:300
SET BATCH-STREAM 1 TIME-LIMITS 0:600
SET BATCH-STREAM 2 TIME-LIMITS 600:6000
SET BATCH-STREAM 3 TIME-LIMITS 0:6000
SET BATCH-STREAM 3 PRIORITY 20:63
START BATCH-STREAM 0:3
is the system.cmd i'm using (stolen from TWONKY)
RUN BATCON seems to not start. It might need started earlier though.
And something like this to enable remote DECnet file access:
SET FAL-STREAM 0 NETWORK DECNET
SET FAL-STREAM 1 NETWORK DECNET
SET FAL-STREAM 2 NETWORK DECNET
DEFINE FILE-ACCESS DEFAULT-PPN [377777,377777]
START FAL-STREAM 0:2
Those I WILL add though.
You may want to try these commands at the OPR> prompt before putting them into
the SYS:SYSTEM.CMD file.
Status of MARLEY TOPS-10 MONITOR at 16:39:09 on 15-May-113
Uptime 43:17, 114% Null time = 114% Idle + 0% Lost, 1% Overhead
More than 100% is not very good. It happened to me with a monitor that was
unable to keep time: try some DAYTIME commands in a couple of minutes and
check if it says the correct wall clock time or not...
It's off by about 30 seconds to a minute...not unexpected due to manually setting the time.
9 [OPR] 751 5 ^C 0
This does not seem very good too: it's not detached, it's running nothing, and
it's in "^C" state... Maybe it's a stale BATCON process?
Yeah, that's a stale job.
HTH,
G.
--
Cory Smelosky
http://gewt.net/ Personal stuff
http://gimme-sympathy.org Experiments
On Wed, 15 May 2013 20:40:01 -0000, you wrote:
Disregard, I made some changed to accounting and everything seems
operational now. I can't SET KSYS, but every single other thing works!
KSYS requires BATCON, so maybe there is something wrong with that. Do you have
tweaked your SYS:SYSTEM.CMD file and set the right permissions?
It should contain at least something like this to enable batch processing:
SET BATCH-STREAM 0 TIME-LIMIT 0:10
SET BATCH-STREAM 1 TIME-LIMIT 0:60
SET BATCH-STREAM 2 TIME-LIMIT 10:100000
START BATCH-STREAM 0:2
And something like this to enable remote DECnet file access:
SET FAL-STREAM 0 NETWORK DECNET
SET FAL-STREAM 1 NETWORK DECNET
SET FAL-STREAM 2 NETWORK DECNET
DEFINE FILE-ACCESS DEFAULT-PPN [377777,377777]
START FAL-STREAM 0:2
You may want to try these commands at the OPR> prompt before putting them into
the SYS:SYSTEM.CMD file.
Status of MARLEY TOPS-10 MONITOR at 16:39:09 on 15-May-113
Uptime 43:17, 114% Null time = 114% Idle + 0% Lost, 1% Overhead
More than 100% is not very good. It happened to me with a monitor that was
unable to keep time: try some DAYTIME commands in a couple of minutes and
check if it says the correct wall clock time or not...
9 [OPR] 751 5 ^C 0
This does not seem very good too: it's not detached, it's running nothing, and
it's in "^C" state... Maybe it's a stale BATCON process?
HTH,
G.
On Wed, 15 May 2013 22:38:35 +0200, you wrote:
There is a section in SIG.MEM that explains very clearly and easily how to
create a complete system backup on tape.
To Cory:
I reply to myself because I remembered an important thing: the TOPS-10 V7.05
monitor has a bug that prevents it from initializing new structures when doing
a scratch install. So, if you create a recovery tape containing that monitor,
you'll not be able to initialize disks and proceed with the installation.
The solution I found is a simple patch that adds a single instruction in a
monitor module which got modified during the very last development cycle and
was probably never throughly tested. When you'll be ready and with some spare
time, remember me to tell you how and what to do! :)
Oh, and there is the patch to enable the idle feature of KLH10 too!
G.
