Do you regularly rebooting your server?

I would like to know if you reboot your SQL Server on a regular basis and why...

I do not regularly reboot our SQL Servers. They do get reboots when patches adn service packs are applied, but past that, I have found no need to reboot the servers.
|||Not by choice but monthly due to patches|||No. Only by office power cutoff (occurs time to time).|||

Never :-)

|||

If you don′t wanna have your data cache and your execution plans dropped every time you boot up the server, don′t do it. :-)

HTH, Jens Suessmeyer.

http://www.sqlserver2005.de

|||If there weren't so dang many OS security patches, we would choose never. But, our company has opted for applying patches every other month. Many of our servers are using WSUS and unattended installs/reboots. We've had no troubles with this.
|||Come on, don't be shy, no one will judge you here. Or you could email me at dr_remove_sql@.hotmail.com if you don't want to tell the world.|||

Louis Davidson - SQL Server MVP wrote:

I would like to know if you reboot your SQL Server on a regular basis and why...

Funny you should ask...

Actually about a week ago our SQL 2005 server began to require a reboot about once per day. I'm no DB engineer or sys engineer so I can't tell you why. Our co-host is looking at he issue for us. I can tell you that there did not seem to be any irregularities in the logs and there's no storage issues. The websites just go down because the service fails, and I reboot it.

Drew

|||

We have our SQL server Servers rebooted every weekend if possible.

Why :

Patches

Some of our older SQL Servers are still on NT and SQL7 due to legacy application.

Development servers run applications that can have memory leaks.

If problems occure on reboot you will have a datum from when the server last rebooted successfully. Although servers are secured you can never be sure that someone has not been installing and not finished of with a reboot to ensure all that was working still works and no other errors have been introduced.

Prevents the 'worry' when a reboot is required and no ones done one for ages - we get this 'worry' with the unix server guys mostly. Rebooting apparently once introduced disk error problems due to power off on.

Cluster Servers - when failing over manually and rebooting the target before the switch ensures the servers are fully functioning and able to reboot. We have had problems with reboots in the past. -normally down to changes that have been made.

|||Hi, everybody.

Opened up a poll for this thread on http://www.sqlserver2005.de/Polls

HTH, Jens Suessmeyer.

http://www.sqlserver2005.de
|||Awesome. Can't wait to see the results :)