| Title: | DEChub/HUBwatch/PROBEwatch CONFERENCE |
| Notice: | Firmware -2, Doc -3, Power -4, HW kits -5, firm load -6&7 |
| Moderator: | NETCAD::COLELLA DT |
| Created: | Wed Nov 13 1991 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 4455 |
| Total number of notes: | 16761 |
I would like to notify two problems that I met with HUBWATCH 3.1 for OSF/1: 1. Crash when disconnecting a concentrator - the configuration used is in dual homing (PHY A and PHY B ports are connected to different concentrators in 2 different hubs). - Hubwatch polling is enabled and set to 10s - 1 concentrator is removed from its slot (it is not the slot used for in Band Management) - we wait for the display by Hubwatch of the concentrator loss (on the physical view) - after display of the concentrator loss, the same concentrator is put back in the same slot. - the initialisation phase of the concentrator takes place (about 45s), then after a rather long delay (about 1 mn), Hubwatch displays the concentrator back to operational in the physical view and crashes immediately after (about 1s) 2. If a power supply is turned off but left in hub, Hubwatch indicates that it is still OK. The faillure is displayed only when the power supply is physically removed from the slot. Are those problems known? are they corrected by patch or release? Thanks, St�phan.
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 1893.1 | Send us the core | NETCAD::FORINO | Mon Jan 16 1995 09:10 | 11 | |
For the crash please read notes 1871.3 and 1871.5 about how to get
crash information and core dumps to us.
For the power supply the HUB will still detect that the power supply is
in the slot even if you turn the switch off. If you look at the help
for the Environmental View, Table 5-1 explains the LED states for the
power supply. I believe that the LED state should go from green to
yellow when you set the switch off. However, you should refresh the
screen to see the state update.
John
| |||||
| 1893.2 | HUBwatch for OSF/1 problem | UTRTSC::GROOT_R | Ronald de Groot | Tue Jan 17 1995 09:11 | 33 |
Hello,
I have a customer with the same problem. The only difference is that
my customer say that HUBwatch crashed 1 sec. after HUBwatch polled
the HUB900 and see the concentrator loss. The software they used is
OSF/1 3.0, HUBwatch 3.1, HUB900 NAM 3.1. and concentrator SW 2.8.
Polling is also set to 10sec (min. value) I have also the crash
info and core dump.
dbx version 3.11.6
Type 'help' for help.
Core file created by program "hubwatch"
signal Segmentation fault at [strcat:136 ,0x3ff800d4b24]
Source not available
(dbx) where
> 0 strcat(0x11fffeea8, 0x140084af8, 0xfffa5a5afffa5a5a, 0x11fffeeae,
0x11fffeba0) ["../../../../
../src/usr/ccs/lib/libc/alpha/strcat.s":136, 0x3ff800d4b24]
1 hub_snmp_oid2str(0x1402132a0, 0x11fffef18, 0xfffa5a5afffa5a5a,
0xfffa5a5afffa5a5a, 0xfffa5a5a
fffa5a5a) ["hub_snmp_ifc.c":3596, 0x120445438]
2 hub_snmp_build_get(0x303037332d2e3638, 0x303037332d2e3638,
0x303037332d2e3638, 0x303037332d2e
3638, 0x303037332d2e3638) ["hub_snmp_ifc.c":2698, 0x1204436e4]
3 hub_req_send(0x303037332d2e3638, 0x303037332d2e3638, 0x303037332d2e3638,
0x303037332d2e3638, 0x303037332d2e3638) ["hub_mgmt_ifc.c":2879, 0x12043c1a8]
(dbx)
Is this a known problem / bug?
Ronald de Groot
| |||||
| 1893.3 | Not a known problem | NETCAD::FORINO | Tue Jan 17 1995 12:36 | 3 | |
This was not a known problem until now.
John
| |||||
| 1893.4 | UTRTSC::GROOT_R | Ronald de Groot | Wed Jan 18 1995 05:01 | 5 | |
John,
Must I open a IPMT case for this problem?
Ronald
| |||||
| 1893.5 | Can do it here | NETCAD::FORINO | Wed Jan 18 1995 12:39 | 3 | |
Not really - we can track it through here if you want.
John
| |||||
| 1893.6 | UTRTSC::GROOT_R | Ronald de Groot | Thu Jan 19 1995 06:12 | 8 | |
John,
>Not really - we can track it through here if you want.
Yes,
Ronald
| |||||
| 1893.7 | UTRTSC::GROOT_R | Ronald de Groot | Tue Jan 31 1995 05:41 | 1 | |
Any status? | |||||
| 1893.8 | SLINK::HOOD | I'd rather be at the Penobscot | Tue Jan 31 1995 11:59 | 8 | |
Do not expect a patch as a result of a bug report in this conference. The best that will happen here is that the next version of HUBwatch may contain the fix. If you have an immediate need, you need to follow the usual channels. Tom Hood HUBwatch | |||||
| 1893.9 | repl on .8 | UTRTSC::GROOT_R | Ronald de Groot | Wed Feb 01 1995 03:43 | 8 |
Tom,
>If you have an immediate need, you need to follow the usual
>channels.
Thats what I asked in .3
Ronald
| |||||
| 1893.10 | You'd be best to log an IPMT case....... | NETCAD::BATTERSBY | Wed Feb 01 1995 08:57 | 8 | |
Ron - You'd be best to open an IPMT case, just so there is a
formal history of the problem in case it is seen by other sites.
It also provides a mechanism for closure on problems. IE: It
is documented as having occured and you have a way of tracking
it too. It keeps the system honest, and provides means to measure
product quality and customer satisfaction criteria.
Bob
| |||||