| Title: | DIGITAL UNIX (FORMERLY KNOWN AS DEC OSF/1) |
| Notice: | Welcome to the Digital UNIX Conference |
| Moderator: | SMURF::DENHAM |
| Created: | Thu Mar 16 1995 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 10068 |
| Total number of notes: | 35879 |
referencing note 8853, where the solution was using a BN24F cable
instead of the cat 5 twisted pair cable, I'm working on a somewhat
similar problem, and I had a couple of questions on it.
Customer is at 3.2g trying to configure a 21140-aa (defpa-aa?) without
much luck.
He's tried 2 controllers on this one system but the results are the
same. The controller comes up, no link errors, link light is ON, no
output errors, output packets as expected, the Atlantic hub he is
connected to indicates that it sees the output packets, but there
are NO input packets. Everything outbound fails obviously.
If the customer takes the cable, moves it over to a VMS system (leaving
the other end in the hub) that comes up fine.
We tried full duplex, half duplex, 100mb half and full....they all
give output errors except the 10mb half duplex which looks like it
SHOULD be working just fine!!!!
Since he is currently using the 4 pair cable, and note 8853 stated that
using a 2 pair cable resolved that issue, I am having him try the 2
pair cable.
Since the hub doesn't seem to mind the 4 pair cable (works to the vms
node) that would mean that it's the de500-aa module that doesn't like it.
Is this a 'known' issue?? Is this documented anywhere? Has anyone else
got one working with a 4 pair cable??
thanks!
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 9283.1 | netrix.lkg.dec.com::thomas | The Code Warrior | Tue Mar 25 1997 16:18 | 2 | |
Did you install the required OSF patch for the DE500-AA? | |||||
| 9283.2 | ?? | CSC32::PITT | Thu Mar 27 1997 09:04 | 42 | |
The patch for the de500-aa for 3.2g says:
PROBLEM: (47944) (Patch ID: OSF375-048)
********
A bug in the DECchip 21140A device used on the DE500-AA controller
causes
reporting of bad incoming packets by the device as being without error.
In most situations, the packet will be discarded by higher-level
protocols,
such as TCP/IP, due to checksum failures. However, this is not always
guaranteed.
The network interface may hang due to the receive process coming to a
halt
within the DE500-AA controller. However, no specific symptoms can be
predicted if and when the data-corruption occurs. It should be noted
that
the bug has only been observed during heavy receive stress testing, and
is
most likely to occur when the 21140A has poor write access to memory
through
the host bridge.
In this patch, the driver recognizes the condition that a bad packet
may
have been delivered as good one by the device, and takes the
appropriate
action (ensures packet is not propogated further into the protocol
layers).
----
This is the only de500 patch I can find for 3.2g. Is there another
patch you're talking about or does this patch fix more than described
above? In this customer's case, there are NO packets coming in.
Plenty of outgoing, nothing incoming.
thanks (I'm going to get him this patch anyways...just cause YOU said
so. 8-}
| |||||
| 9283.3 | DE500-AA or DEFPA??? | SMURF::GILLUM | Kirt Gillum | Thu Mar 27 1997 14:49 | 7 |
Exactly what kind of device are you worried about here? The DE500-AA
or the DEFPA-AA?
The DE500-AA is Ethernet and the DEFPA is FDDI. FDDI adapters can't
talk to Ethernet repeaters.
| |||||