🚀 go-pugleaf

install one of the logging tools and start reading , it sounds like
something is not releasing memory , in which case the machine will
reboot  auto magically ,  use a memory manager like fixit (your choice)
and tag the logs for reading until you do actually diagnose the leak,

Simon wrote:
> 
> But there are events messagees immediately prior to the reboot
> suggesting that the system is out of memory.
> 
> Damien McBain wrote:
> 
> > Simon wrote:
> >
> >>I have a server with Xeon 2.4Ghz processor, 4GB RAM, 60GB HDD and it
> >>runs no programs, only 2 additional services.
> >>
> >>The 2 services related to Text-to-speech software called Speechify.
> >>
> >>Over the past couple of months the server has been rebooting itself
> >>every few days.
> >>
> >>In the Event Viewer the sequence of events leading to an unplanned
> >>system reboot starts with an error with Event ID 2019. This error
> >>under Windows 2000 Server is documented at
> >>http://www.jsiinc.com/SUBO/tip7100/rh7141.htm - this problem can occur
> >>if you have an Intel network adapter and you do NOT have the latest
> >>driver. "Previous drivers leaked memory, causing the subject behavior.
> >>Subsequent errors in the Event Viewer related to insufficient memory
> >>resources being available", despite a physical memory resource of 4GB,
> >>which indicates to me a severe memory leak.
> >>
> >>The latest drivers for both the 100Mb and 1000Mb Intel NIC's (standard
> >>on the server) were updated with drivers dated December 2003, where
> >>the previous drivers were dated March 2003.
> >>
> >>However, this has NOT solved the problem and the server continues to
> >>reboot randomly. There is no pattern to the timing of the reboots,
> >>except that they are every 4-7 days.
> >>
> >>I am thinking of going down the track of a memory management tool for
> >>the machine. Any suggestions on which one? Any other suggested courses
> >>of action?
> >
> >
> > No experience with *real* servers but I had the same problem with another
> > machine and the PSU turned out to be the culprit. It would power off,
> > usually when accesing network resources via explorer (I really never
> > bothered to track exactly what) and restart.
> >
> >