Hi there, our main Border Manager Server always hangs on unloading
proxy.nlm if it has been up for more than one work day.

The only way I can bring the server down is force deactivate all the
NSS Pools (All Cache volumes are traditional) and then enter the
debugger and quit out to dos.

It is a NetWare 6 server with SP3 and BorderManager 3.7 with SP2 on anew HP Proliant ML530 G2.

Border Manager is only being used for proxy for about 500 users -no
filtering, access control or VPN.

The server is a ICP Cache Server for three other Border Manager 3.7
Cache Hierarachy Clients with 100 users between all three. They are at

a similar software patch level and have never had a hang on unloadingproxy.nlm.

Software Config:

Patches applied include:

bm37fp3f
tcp607k
edir862sp4
httpstk3
nls603ft
nw56up3
nwftpd9
nw6nss3c

I am using Craig's (thank you for your excellent manual) proxy.cfg and

have the lines:

RES_BadAddressListCleanUp=1
DoNotSaveMemoryCacheDuringUnload=1

in there as well.

There are 4 x 4GB cache volumes with 16KB blocks & no sub-alloc or
compression.

The other few NSS volumes do not run compression.

The HP Prolaint Support Pack (PSP) is version 6.40

Hardware Config:

HP Proliant ML530 G2 with Dual 2.8GHz Xeon CPU's
3GB RAM
6 x 36GB Disk RAID 1 + 0 array being driven by Smart Array 6402
HP Broadcom Gigabit NIC


I am at my wits end with this one. I have already reinstalled the
whole server once. The server is tuned according to Craig's excellentmanual.


Has anyone encountered issues similar to this?

Cheers, Andy.

==
Andy Simpson CNE MCSE