Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
sqlguy7777
Starting Member
14 Posts |
 Posted - 2011-09-06 : 21:23:05
|
| Howdy,We are running SQL 2008 R2 10.50.1790 ( RTM + KB2494086 ) on Windows 2008 R2.Just wondering if this is normal - we see occasional database DBCC CHECKDB and db restarts in the error log.To me it doesnt look right.We were getting :"The log for the database XYZ is not available""Error 9001, Severity 21, State 1"But now we are not getting that - we are just getting what appears to be regular restarts of most of the databases on the server.A typical sequence over approx 2 minutes is this:"Starting up database XYZ""CHECKDB for database XYZ finished without errors""Starting up database XYZ""CHECKDB for database XYZ finished without errors""Starting up database XYZ""CHECKDB for database XYZ finished without errors""Starting up database XYZ""CHECKDB for database XYZ finished without errors""Starting up database XYZ""CHECKDB for database XYZ finished without errors"We also notice databases randomly ( briefly ) displaying "In Recovery" and then coming good ( the "In Recovery" message disappears ).Logic says this just cant be a good thing - it happens to databases randomly and the server is a VM hosted on a powerful ESX host with fibre channel connection to a large HDS SAN.To me, it looks like SQL briefly loses connection to the disk.When I run a big query against a random database ( e.g. database ABC ) that will give the server a good workout, occasionally we will see this :"Starting up database XYZ""CHECKDB for database XYZ finished without errors""Starting up database XYZ""CHECKDB for database XYZ finished without errors" Its almost as if SQL is being overwhelmed by load and being flaky - even though the CPU is only about 50%This happens on both VM connected disk and local hard disk on a PC.It happens with SQL at 10.50.1600 ( on PC ) and 10.50.1790 ( VM with HBA fibre connected disk to ESX host ). Next trick will be to try SP1.Thoughts please? |
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
Kristen
Test
22859 Posts |
Posted - 2011-09-07 : 03:11:28
|
| I think Tara's is most likely explanation.Auto Close used to be (still is?) the default on MSDE / "stand alone" versions of SQL server, so databases that started life there and have been migrated to full-fat version of SQL Server are more likely to have that setting. |
 |
|
|
GilaMonster
Master Smack Fu Yak Hacker
4507 Posts |
Posted - 2011-09-07 : 04:01:57
|
| Very likely autoclose.Those checkDB messages aren't SQL running checkDB. It's saying when checkDB last ran (read the message carefully)--Gail ShawSQL Server MVP |
|
|
|
russell
Pyro-ma-ni-yak
5072 Posts |
Posted - 2011-09-07 : 21:44:52
|
| That or reboots |
|
|
|
sqlguy7777
Starting Member
14 Posts |
Posted - 2011-09-07 : 21:49:44
|
| HiYep that was it. Performance is up to, as I guess you'd expect.Much appreciated. |
|
|
|
tkizer
Almighty SQL Goddess
38200 Posts |
|
|
|
|
|
Seeing random database restarts
Tara Kizer