Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: (ASCEND) IP address leaks
At 04:39 AM 7/21/98 -0400, Rob Marlowe wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>We're currently running 6.1.0 tik.m40 on our Max 4048's. The first
>Max in my hunt group is leaking IP addresses pretty bad. Is there
>anything I can do to make things better besides giving it plenty of
>extra IP addresses and rebooting it every 48 hours or so?
As you will see other replies to this, I'll just add a few comments.
When 6.1.0 was released, it had been run on quite a few customer sites
in BETA. Unfortunately, even "quite a few" does not cover all of the possible
configuration combinations of a complex box like the MAX - and we found out
with the release that some sites would get IP address leaks and possible
out-of-memory conditions (including resets, failed RADIUS requests etc).
6.1.3 was released to address the memory issues and fixed what we thought
was the cause of the IP address leaks. We confirmed this by running it on
customer sites that had been reporting the IP leaks on 6.1.0 - all seemed to
be fixed. But once again, it seems we didn't get all of the possible
configuration combinations that could cause the leak and still some people
get the leak in 6.1.3.
We appear to have isolated it to being related to consistent authentication
or MPP failure when bringing up the second channel. One workaround that seems
to help (the reason being quite complex :) is to set the parameter:
ethernet/Mod Config/Auth/Auth Pool=No
We should have a further fix in the next point release possibly this week. It
has been confirmed on some customer sites, but we are trying to verify all
known configurations that may cause the leak.
Kevin Smith Ascend Communications...
Customer Satisfaction ...where Network Solutions never end.
++ Ascend Users Mailing List ++
To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd: <http://www.nealis.net/ascend/faq>
Follow-Ups:
References: