High utilization on NW3.12 server
I just upgraded our NW3.12 server from a P3-500 w/128 MB to a P4-2.8
w/512 ECC. The old hard drives were 9 gig and the new hard drives are
80 gigs. There are three volumes on the server that were extended to
25-35 gigs each. The sys volume is 500 MB. The short term memory was
increased to 32 MB.
The old server typically has 50 open connections w/ roughly 175 open
files. It rarely hits 90% utilization. The new server hits 90+%
utilization any time files are copied. Additionally, when copying
files several work stations lose their network connection to the
server. I've also noticed that the server console is frozen and the
HDD light is on solid while files are being copied.
Does anyone know why the faster machine is running slower than the
original machine and why the utilization is so high when copying
files? Are there additional parameters I should change? Here's some
My total server memory is 535,122,872
Here's what I have:
Original cache buffers: 126,659
Total cache buffers: 117,819
Packet receive buffers: 100 (although set to a max of 1000)
Directory cache buffers: 35
Permanent memory pool: 987,500 - 0% w/ 989,940 in use.
Alloc memory pool: 20,011,212 - 4% 12,815,824 is in use.
Cache buffers: 497,646,336 - 93%
Cache movable memory: 14,707,968 - 3%
Cache non-movable memory: 1,769,856 - 0%
Do these look correct or does something need adjusting?
Re: High utilization on NW3.12 server
On 20 Aug 2003 04:23:24 -0700, [email]email@example.com[/email] (thdrman) wrote:
>Do these look correct or does something need adjusting?[/color]
Suggest you increase the value of the SET parameter:
SET MAXIMUM CONCURRENT DISK CACHE WRITES=1000
(This affects the size of "bucket" of data which is written to disk in
Some more explanation here: