| T.R | Title | User | Personal
Name | Date | Lines |
|---|
| 2090.1 | more information | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Mon Jan 13 1992 16:45 | 91 |
|
More info. Here are the results when run from the command line.
*** Spanning Tree Map FM (STM_FM) T1.2.4 Messages ***
(I) Generating spanning tree map for domain spanning_tree_map.
(I) Building autoregistration command/script file...
(I) No registered bridges found in the mcc registry.
(I) Created Autoregistration command/script file:
mcc_stm_spanning_tree_map_aut
oreg.com
(I) Polling bridges for spanning tree attributes...
(I) Building logical spanning tree topology map...
(I) Writing topology map file...
ERROR2!!
(E) Error opening map file for domain: spanning_tree_map
%SYSTEM-F-ACCVIO, access violation, reason mask=00, virtual
address=0000001D, PC
=0000B427, PSL=03C00000
%TRACE-F-TRACEBACK, symbolic stack dump follows
module name routine name line rel PC
abs PC
0000B427
0000B427
004B5625
004B5625
004B3DA5
004B3DA5
004B399C
004B399C
004AFAA1
004AFAA1
004AF44C
004AF44C
000AA795
000AA795
00083FC1
00083FC1
0007E9B8
0007E9B8
0008DFB1
0008DFB1
000AA89D
000AA89D
000F1B2C
000F1B2C
000B3483
000B3483
000B31E3
000B31E3
000B3139
000B3139
000B2B7B
000B2B7B
000B262C
000B262C
000B253B
000B253B
001C705C
001C705C
001C7D40
001C7D40
001C6871
001C6871
001C9861
001C9861
001C8C36
001C8C36
001C5D3C
001C5D3C
001C4584
001C4584
000C6E9A
000C6E9A
MCC_MAIN main 8258 00000208
00001E3C
000056F7
000056F7
MCC_KERNEL_INIT MCC_KERNEL_INITIALIZE 163 00000027
0000572B
000056D2
000056D2
|
| 2090.2 | Domain must be created before running stm_fm | QUIVER::HAROKOPUS | | Tue Jan 14 1992 10:47 | 15 |
| My guess is that the domain spanning_tree_map does not exist and this is why
you are having this problem. The IMPM uses the in domain qualifier to pass
in the domain name. If no domain is opened then the stm fm is using a
default domain name (spanning_tree_map). However, it does not create this
domain. This is a problem that I plan to fix.
If you want to run stm fm using FCL use the in domain qualifier and specify
a domain that exists. i.e.
mcc>create domain tree_map
mcc>automap mcc 0 stm_fm autoregister=true, in domain=tree_map.
Sorry for the confusion.
Bob
|
| 2090.3 | domain exists in impm scenario | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Tue Jan 14 1992 11:15 | 15 |
|
Hi Bob,
Thanks for the info. In my .1 you are right. I did not specify a domain and
I remember reading that in the release notes of autopology guide.
However, when I was in the impm i started in a domain called cxn1, created
a sub domain called bridges and had moved into that domain before started the
autotopology process. Is there something else I need to do there to make
sure that the proper domain is specified? I did notice that even when I was
in domain bridges the map files were still created with the cxn1 domain name.
Thanks,
Scott
|
| 2090.4 | That's a known bug | QUIVER::HAROKOPUS | | Tue Jan 14 1992 11:41 | 14 |
| Hi Scott,
That's a known bug with the LAN Autotopology PM. It is using the root
domain parameter instead of the in domain qualifier (I lied before) to get
the domain name from the IMPM. Therefore, hierarchiccal domains will not
work properly as it will always use the root domain (as you have seen).
The only work around is open each domain explicitly from the File menu
and make sure that the there are no domains above in the hierarchy.
Bob
P.S. How does your map look?
|
| 2090.5 | still no map | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Tue Jan 14 1992 12:20 | 13 |
|
Bob,
I don't think that I'm running into domain problems any longer.
I removed the bridges subdomain from under cxn1 and used the file pull
down to create a new domain. The domain name is .mcc.domain.bridges and
there are no domains above it. I still get the same access violation
information from within the impm. When I run the automap command from
the fcl I also get the access violation but no longer see the "error2"
message. Any other ideas?
Thanks,
Scott
|
| 2090.6 | Need more info | QUIVER::HAROKOPUS | | Tue Jan 14 1992 13:38 | 11 |
| Hi Scott,
The access violation is occuring when the map file is being built. Is there
a map file being created by the stm fm before it dies or is it still
failing on the map file open?
Thanks,
Bob
|
| 2090.7 | requested info | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Tue Jan 14 1992 15:48 | 34 |
|
Bob,
A new map file is being created, although not where mcc_maps is
pointing. Once the access violation occurs the impm is dead so I
control_y out at the decterm. Here is the contents of the map file.
Directory SYS$SYSROOT:[SYSMGR]
MCC_2394000400AA000001CAD286CF59CD20.MCC_MAP_BRIDGES;1
1/3
Total of 1 file, 1/3 blocks.
$ ty MCC_2394000400AA000001CAD286CF59CD20.MCC_MAP_BRIDGES;1
version 0
domain DSP_NS:.mcc.domain.bridges
last_x 0.975000 last_y -0.333000
object_type 7
text Spanning Tree Root: 08-00-2B-0F-B1-AE
x 1.300000 y 0.449550
object_type 7
text Spanning Tree Mode: LAN Bridge 100 Mode
x 1.300000 y 0.416250
Along a different line...
While reading the release notes I saw a section that says that a second
mcc window is opened while the automap is running. I don't know what it
is worth but the only time I have seen this happen is when I purposely
tried to start the autotopology without starting the elm am first.
Thanks,
Scott
|
| 2090.8 | the root bridge was the key | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Tue Jan 14 1992 19:02 | 16 |
|
Bob,
You were right. The segment I was running the listener on is connected
to the rest of the building network via a bridge that is filtering
sysid messages, and therefore I never find the root bridge. I
reconfigured my segment a little so that I created a root bridge that
I could find. I didn't end up with any complete maps because of
my configuration, but did get by without the access violation.
I'll expermient further when I can create a more involved private lan
to work with.
Thanks for all the help.
Scott
|
| 2090.9 | | QUIVER::HAROKOPUS | | Wed Jan 15 1992 09:43 | 6 |
| Thanks Scott,
I'll be working on the error handling so that acc-vio won't occur in
the future. Keep me posted on how it goes.
Bob
|
| 2090.10 | a little strange behaviour | NSCRUE::KEANE | U.S. NIS Service Development and Delivery | Fri Jan 17 1992 14:16 | 20 |
|
Bob,
Thought you might be interested in this.
I finally got several lan bridges set up in a private lan where I could
make sure that the listener found the root bridge. After checking to see that
the listener knew about all the bridges I tried to start the autotopoloy from
the impm and got the following error:
Map window message
Dispatch local management module file access error during probe
I then went to the fcl and issued the automap command and it went through a
successful map build. After it worked correctly from fcl the failure from the
impm cleared up.
Scott
|
| 2090.11 | I'll see if I can reproduce it | QUIVER::HAROKOPUS | | Mon Jan 20 1992 10:21 | 7 |
| Scott,
I've never seen that problem before, but I'll look into it.
Thanks,
Bob
|