Posted by Greg D. Moore \(Strider\) on 12/06/07 22:50

"Neil" <nospam@nospam.net> wrote in message
news:NQZ5j.69213$RX.4340@newssvr11.news.prodigy.net...
> OK, that's fine. But my POV is that rebooting in general, and resetting
> the system memory, is a good thing to do. I take it that you disagree?
>

Yes. I disagree.

I've run systems with uptimes measured in years.

I see no value in rebooting a correctly setup system.


> "Michael Abair" <mabair@autotask.com> wrote in message
> news:eTxBHcEOIHA.2064@TK2MSFTNGP06.phx.gbl...
>> Just because the reboot fixed the issue doesn't mean that a preventative
>> reboot will stop it from occuring.
>>
>> A reboot might not have been needed to resolve the issue in the first
>> place. Possibly a configuration change
>> or bouncing a service could have resolved it. My advice is to spend some
>> time researching the issue and if
>> you cant find a way to resolve it then find a way to detect it and have
>> it perform the least intrusive form of corrective action.
>>
>> -Mike
>>
>> "Neil" <nospam@nospam.net> wrote in message
>> news:5CY5j.76025$YL5.36846@newssvr29.news.prodigy.net...
>>> Running a SQL 7 system on a Windows 2000 server using Access 2000 on
>>> client machines as a front end. System administrator currently reboots
>>> the server once a month. Yesterday we had some weird thing with the
>>> database where users were getting ODBC errors when trying to access it.
>>> Rebooted the server, everything was fine. Suggested to the sa that he
>>> reboot the server once a week. He said he already does it once a month,
>>> and that's sufficient.
>>>
>>> My POV is that: a) doing it once a week might prevent situations such as
>>> the one yesterday; b) even without situations like the one yesterday,
>>> performance may be degrading over the course of the month, without our
>>> being aware of it, and rebooting once a week might help performance.
>>>
>>> Any thoughts on this?
>>>
>>> Thanks!
>>>
>>> Neil
>>>
>>
>>
>
>

[Back to original message]