any comments on the statement: "a nightly server reboot will lead to quick hardware failure

[ams2013mrb]

My company uses a third party for technical support of our servers and we are having some issue with their technical support performance. We are running older hardware (2007/2008) and plan to replace them later this year.  We plan to replace the servers in about 4 to 6 months but this support company has been pushing sales for the past year and almost always has the solution of sales of buying new equipment or software from them to correct the problem.

We were told today that our server problem is caused by a Windows kernel memory leak and that server needs to be replaced. I asked for a singly nightly reboot of the server to correct the memory leak problem and was told that this will only lead to hardware failure and was only used in the last millenium as a potential fix to a memory leak.  Anyone have comments about this: that a daily server reboot will cause hardware failure or know of a whitepaper from HP that says to minimize server restarts to increase hardware life?

[Torsten. · Accepted solution]

nonsense - IMHO.

[ams2013mrb]

Thanks. Was Mine also.

[Robert_Jewell]

Agreed...its bull.  A reboot doesnt cause the server hardware to do anything it doesnt normally do, except maybe run the POST routines during the reset.

Now, if you are shutting down and powering the server off and on each time, then perhaps you are adding wear and tear to the power supplies and disk drives.  However, even this is debatable.

-Bob