> just to let you know. For me, all 2.1.xx kernels are hanging trying
> to allocate TCP buffers (I reported this problem), and I tried vhand
> patch in the hope that it may help, as it deals with memory management.
It's still in beta :)
>
> Unfortunately, it did not. I got the system hung with the same
> symptoms in less then 24 hours uptime. Also, perfomance of cpu
> intensive applications (mpg123) dropped noticably with vhand.
I'm working on that one, and it seems like Joe Fouch has given
me the 'golden hint' on what to do. So I implemented his idea
and tested... It improved performance in CPU intensive applications,
but I/O intensive stuff really suffers.
In vhand-2.1.66 this should be better...
>
> I have a 120MHz 486dx4 w/16Mb, kernel 2.1.63+vhand-2.1.63.
>
> BTW, Zlatko's patch seemed to cure the hang problem, but after a few
> days, the system freezed anyway at the same place.
Hmmm, could you give me some more details:
- type of network card
- amount of network-buffers allocated each second (!)
- size of allocated network buffers
- amount of swapping/paging going on
- how was the CPU usage of kswapd/vhand during that time?
thanks and good luck,
Rik.
----------
Send Linux memory-management wishes to me: I'm currently looking
for something to hack...