This message has something to do with exceeding the number of processes. Try
increasing the number of processes to Proc.
60 is the message error number, not the node number! It has something to
do
with FLEET in Net. I will have to speak to the Net guru on Monday.
“Tony Williams” <> twillqnx@yahoo.com> > wrote in message
news:c53v1t$pj6$> 1@inn.qnx.com> …
Hugh,
Can you tell me what the following from netinfo means? Could this be
causing Net to crash? I’ve tried changing the ethernet card on node 60,
but I’m still getting these messages from netinfo on node 1.
12:04:22 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:04:37 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:04:53 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:04:55 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:04:58 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:05:24 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:05:26 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:05:31 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
12:05:34 0 ( 60) Net ( tx) vid was too large (get_tx_vseq)
Hugh,
Could you please tell me what the REPNE MOVSB command is trying to do?
Unfortunately not. I don’t have a map file to see where it is in the
code,
but it looks as though it is in the driver code. In QNX4 all the fleet
protocol is handled in the driver, so yes, it could be a bad message
that is
causing Net to crash, especially if you are seeing a 010101… MAC
address.
\
“Tony Williams” <> twillqnx@yahoo.com> > wrote in message
news:c4v49g$ivk$> 1@inn.qnx.com> …
Hugh,
I thought of that on Friday and moved the server hard disk to
another
computer and Net still crashed. I ran chkfsys on the disk
partitions
and found no errors.
On Sunday I moved all critical server tasks to another computer and
made
it our node 1 to elminate a hard drive issue and the Net task
started
crashing on this new computer. I’ve changed all hardware and the
problem only exists on node 1 on our network.
Today I’ve moved nameloc to run on our node 2 and all computers on
our
network are kicking the licenses from that computer now. The
problem
has not happened on node 2 today, so I feel the problem is not with
nameloc like I mentioned earlier.
Has anyone been able to run a debugger on the dump file I sent? I’m
curious what is going on at 0015:0000A5A9 in the Net task.
Yes, I took a look at the dump and it is at a REPNE MOVSB command
that
it is
crashing. Have you tried replacing the Net on your hard disk?
Thanks,
Tony
Hugh Brown wrote:
OK, have you considered changing the memory on the server? The fact
that
multiple drivers all crash, seems to indicate a hardware problem of
sorts.
If only one driver was crashing, then I would say that it is a
software
problem.
“Tony Williams” <> twillqnx@yahoo.com> > wrote in message
news:c4s9jd$6gn$> 1@inn.qnx.com> …
Hi Hugh,
I’ve tried different cards using different chipsets/drivers and
the
Net
task crashed with all of them. All these types of cards are being
used
on other computers on our network.
The Net task even crashed when I rebooted the computer and did not
start
the Tcpip, so I’m thinking this is a FLEET related problem. When
the
Net task crashes, the ethernet driver continues to run. I was
expecting
the ethernet drive to die too.
Any ideas what the bogus MAC addresses could be coming from? I’ve
attached my Net.dmp file.
Let me know if you find anything.
Thanks,
Tony
Hugh Brown wrote:
When you changed the network card, did you change it with another
Intel
card, or did you try another type of card? Net can crash with a
driver
error
or a Tcpip error, so I’m trying to eliminate which one it is. If
you
try
another type of ethernet adapter and it still crashes, then it is
not
a
driver problem.
Could you also please post your dump file.
Thanks.
“Tony Williams” <> twillqnx@yahoo.com> > wrote in message
news:c4s524$39c$> 1@inn.qnx.com> …
We have a QNX network which consists of approximately 150 nodes.
The
main server for this network starting have the Net task die with
a
SIGSEGV a few days ago. The task dies 0015:0000A5A9 every time
and
a
dump file is produced. The server where the task is crashing is
running
QNX 4.25 and here is the version information:
PROGRAM NAME VERSION DATE
sys/Proc32 Proc 4.25L Feb 15 2001
sys/Proc32 Slib16 4.23G Oct 04 1996
sys/Slib32 Slib32 4.24B Aug 12 1997
/bin/Fsys Fsys32 4.24V Feb 18 2000
/bin/Fsys Floppy 4.24B Aug 19 1997
/bin/Fsys.eide eide 4.25A Feb 09 2000
//1/bin/Dev32 Dev32 4.23G Oct 04 1996
//1/bin/Dev32.ansi Dev32.ansi 4.23H Nov 21 1996
//1/usr/bin/int10 INT10h 4.23B Feb 13 1997
//1/bin/Dev32.ser Dev.ser 4.25A May 24 2002
//1/bin/Dev32.pty Dev32.pty 4.23G Oct 04 1996
//1/bin/Pipe Pipe 4.23A Feb 26 1996
//1/bin/Net Net 4.25C Aug 30 1999
//1/bin/Net.ether82557 Net.ether825 4.25G Jan 11 2001
//1/bin/Mouse Mouse 4.24A Aug 22 1997
//1//5.0/usr/ucb/Tcpip Tcpip 5.00A Jan 26 2001
//1/bin/Iso9660fsys Iso9660fsys 4.23D Mar 20 2000
//1/bin/cron cron 4.23B Oct 30 1997
//1/bin/Mqueue mqueue 4.24A Aug 30 1999
//1//bin/child_adm 4.24A Aug 30 1999
//1/usr/bin/lpsrvr lpsrvr 4.24A Jun 26 1997
Here are the things I tried to eliminate this problem:
- Changed ethernet cards.
- Changed switch ports.
- Changed computers.
- Shutdown inetd and non-critical tasks.
No mater what I try, the computer running nameloc seems to have
the
Net
task SIGSEGV. If I unplug the ethernet cable, the Net task does
not
SIGSEGV.
Another strange thing I’m seeing is an ethernet packet with a
source
and
destination address of 010101 010101. Any ideas what that could
be?
I
see these packets occasionally using netinfo and netsniff.
Any help would be appreciated.