Home | Contact Us | FAQ | Search & Site Map | Link to Us
Sign In | Join | Other 45 Sites in Network
Home
Discussion Groups
DB Engine
SQL ServerMSDESQL Server CE
Services
Analysis (Data Mining)Analysis (OLAP)DTSIntegration ServicesNotification ServicesReporting Services
Programming
CLRConnectivitySQLXML
Other Technologies
ClusteringEnglish QueryFull-Text SearchReplicationService Broker
General
Data WarehousingPerformanceSecuritySetupSQL Server ToolsOther SQL Server Topics
DirectoryUser Groups
Related Topics
MS AccessOther DB ProductsMS Server Products.NET DevelopmentVB DevelopmentJava DevelopmentMore Topics ...

SQL Server Forum / Other Technologies / Clustering / September 2006

Tip: Looking for answers? Try searching our database.

Cluster Failover

Thread view: 
Enable EMail Alerts  Start New Thread
Thread rating: 
Scopus69 - 22 Sep 2006 23:22 GMT
We have a 3rd party vendor application that runs on IIS and connects to a
back end SQL Cluster.  Sometimes a user will try and run a commission report
and the SQL server takes a big hit on the CPU's (Dual 2.0 Ghz Xeon's) and I
will get a error in MOM 2005 that the  "Server Performance Thresholds
SQLSERVR Process > 90% CPU for 15 minutes"   Of course the end user will try
and run it two or three more times taking up more resources.

If I do a fail over to the other SQL server in the cluster the CPU's will
drop down to their normal ranges and everything works fine.

My question is when performing the "failover" that 10 - 15 seconds it takes
to do this what happens to any data that is trying to write to the database?  
Is it lost? And can this cause corruption in the database itself?

Thanks!
Tom Moreau - 22 Sep 2006 23:30 GMT
Failovers don't cause corruption.  Rather, any active transactions are
rolled back when the backup node takes over from the primary node.

Signature

  Tom

----------------------------------------------------
Thomas A. Moreau, BSc, PhD, MCSE, MCDBA, MCITP, MCTS
SQL Server MVP
Toronto, ON   Canada
.

We have a 3rd party vendor application that runs on IIS and connects to a
back end SQL Cluster.  Sometimes a user will try and run a commission report
and the SQL server takes a big hit on the CPU's (Dual 2.0 Ghz Xeon's) and I
will get a error in MOM 2005 that the  "Server Performance Thresholds
SQLSERVR Process > 90% CPU for 15 minutes"   Of course the end user will try
and run it two or three more times taking up more resources.

If I do a fail over to the other SQL server in the cluster the CPU's will
drop down to their normal ranges and everything works fine.

My question is when performing the "failover" that 10 - 15 seconds it takes
to do this what happens to any data that is trying to write to the database?
Is it lost? And can this cause corruption in the database itself?

Thanks!
Scopus69 - 22 Sep 2006 23:41 GMT
When you mean "rolled back" the transactions are basically held in cache
until the backup server has taken complete control?  Is that correct?

One user received the following error during failover in their web session:

[DBNETLIB][ConnectionRead (recv()).]General network error. Check your
network documentation. in Microsoft OLE DB Provider for SQL Server

Is this a concern or only the period before the failover is complete?

Thanks!  

> Failovers don't cause corruption.  Rather, any active transactions are
> rolled back when the backup node takes over from the primary node.
[quoted text clipped - 14 lines]
>
> Thanks!
Tom Moreau - 23 Sep 2006 01:15 GMT
Not exactly.  Anything in cache evaporates.  Transactions are written to the
transaction log.  If a COMMIT record is not written, then when SQL Server
comes up, all of the work done thus far in that transaction is backed out.
This is true whether you are using a cluster or not.

The network errors go away once SQL Server has come back up again.  Clean
apps will try to reconnect.

Signature

   Tom

----------------------------------------------------
Thomas A. Moreau, BSc, PhD, MCSE, MCDBA
SQL Server MVP
Toronto, ON   Canada

When you mean "rolled back" the transactions are basically held in cache
until the backup server has taken complete control?  Is that correct?

One user received the following error during failover in their web session:

[DBNETLIB][ConnectionRead (recv()).]General network error. Check your
network documentation. in Microsoft OLE DB Provider for SQL Server

Is this a concern or only the period before the failover is complete?

Thanks!

"Tom Moreau" wrote:

> Failovers don't cause corruption.  Rather, any active transactions are
> rolled back when the backup node takes over from the primary node.
[quoted text clipped - 19 lines]
>
> Thanks!
 
Sign In
Join
My Latest Posts
My Monitored Threads
My Blog
My Photo Gallery
My Profile
My Homepage

Start New Thread
Enable EMail Alerts
Rate this Thread



©2009 Advenet LLC   Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.