January 18, 2010 at 8:58 am
Hi all,
Is it true that when running a 2 node a-p cluster, the only way to get Reporting Services installed on the cluster is on the SQL 2008 Enterprise Edition? Our support team have installed Reporting onto only 1 of the nodes in the cluster, which has stopped SQL from running on it. We are currently running SQL 2008 Standard Edition. OS we are running is WS2008 x64 Enterprise Edition.
Thanks
Martyn
January 18, 2010 at 9:10 am
As far as I'm aware Reporting Services is not cluster aware, regardless of what Edition you are running. Though the databases can be installed on Clustered SQL Server instances and will fail over successfully, the Reporting Services service itself is not cluster aware and will need to be installed on all servers in the cluster, plus the service will not fail over automatically, so you'll need to implement something to deal with the failover.
January 19, 2010 at 1:55 am
thanks for that, I need to get a better understanding of how it works and how they have installed it as I'm sure the DB has not been clustered.
Originally we were told we needed a seperate server for reporting, but the revised (cheaper) plan was to simply cluster it....headache
Thanks
January 19, 2010 at 2:25 am
No problem.
You might want to check out this link http://blogs.msdn.com/mariae/archive/2009/04/01/reporting-services-resilience-recommendations.aspx.
There are some good snippets about getting Reporting Services onto a cluster.
January 19, 2010 at 8:19 am
Quick update...
I have installed reporting services onto a stand alone server and clustered the reporting database.
This now works fine
I just hope that I can remove reporting services from node1 in the cluster and that SQL will now fail over to it as at the moment it will not run on it since reporting services was installed!
I do see light at the end of the tunnel finally!
Many thanks
Martyn
January 20, 2010 at 4:51 am
Getting SSRS to install on a cluster is tricky. It is a problem I am still working through and do not have a complete answer.
One way to get it installed is to use a different instance name for the SSRS install to the DB services cluster and the SSAS cluster.
For example, assume you have a Windows cluster called PDGBSQLC01 with nodes PDGBSQLC01N01 and PDGBSQLC01N02. You can install SQL SB and SSAS both with the default instance name MSSQLSERVER, and call the clusters PDGBSQLC01DB and PDGBSQLC01AS. You can then install SSRS with an instance name of (say) PDGBSQLC01RS on both nodes, and configure SSRS to use a RS catalog DB on the SQL cluster PDGBSQLC01DB.
If you try to install SSRS with the default instance name of MSSQLSERVER after you have installed the SQLDB cluster, it will not let you do the SSRS install. If you install SSRS with the default instance name of MSSQLSERVER before you do the SQL DB cluster, it will not let you do the SQL DB cluster install.
I am still working through scenarios where SSRS is installed at the same time as the SQL DB cluster, to see if I can get a working cluster where all of SQL DB, SSAS and SSRS have the default instance name of MSSQLSERVER.
Original author: https://github.com/SQL-FineBuild/Common/wiki/ 1-click install and best practice configuration of SQL Server 2019, 2017 2016, 2014, 2012, 2008 R2, 2008 and 2005.
When I give food to the poor they call me a saint. When I ask why they are poor they call me a communist - Archbishop Hélder Câmara
January 20, 2010 at 4:57 am
thanks for the help.
The main problem here we had was not enough time to test the system before they needed it to go live! We had 4 weeks, then were told it had been shortened to two days. Not my call on this one, powers above...
So we now have a live system in use, so downtime is tricky, so I think to make it easier I will go with stickig reporting on a dedicated VM this time and cluster the reporting DB 🙂
Regards
Martyn
January 20, 2010 at 5:07 am
martyn, your approach sounds like a good approach given the constraints you have.
Overall, my understanding is that most people prefer to put SSRS on dedicated servers, even if they use a RS catalog DB on a 'normal' SQL server box. However, as the number of cores amd memory available on commodity servers increases then the need to split these things out decreases.
You can currently get a 24-core blade with 128GB memory for about GBP 40k, and using this amount of power for a single Windows instance running SQL can save a lot of license and operational costs. I think this scale of change in hardware capabilities will drive a lot of future consolidation work so that people have far fewer SQL images in the future than they felt were needed in the past.
Original author: https://github.com/SQL-FineBuild/Common/wiki/ 1-click install and best practice configuration of SQL Server 2019, 2017 2016, 2014, 2012, 2008 R2, 2008 and 2005.
When I give food to the poor they call me a saint. When I ask why they are poor they call me a communist - Archbishop Hélder Câmara
January 20, 2010 at 8:07 am
2. Install IIS and RS on the default instance, mssqlserver, on both N1 and N2.
3. On the N1 - Configure the RS to create the RS databases on the the cluster and RS server name as V2.
4. On the N2 - with the same RS server name and V2, Configure the RS to connect the previously created ReportServer and ReportServerTempDB.
5. Make sure to use the same Window Service ID on both nodes
6. The Web Service Identity is NT Authority\NetworkService
7. As RS is not the cluster aware, so both of the RS services should be Started
8. When connect to the RS, it will be something like http://v2/reports/
January 21, 2010 at 12:40 am
got to the bottom of the problems this morning...
disabling SQL reporting Services on node1 has now allowed it to accept SQL in the event of a failover. Really not too sure as to why this would stop it, but I am happy it is working again as we needed failover abilities on this live app - yippee...
Also,
SQL cluster can now be accessed via IP address, I enabled TCP/IP and PIPES protocols and rebooted server - these used to be set to enabled but have automatically diasabled themselves when the cluster went wrong
So, I'm a lot happier today! Reporting Services will be left on it's own box going forward.
Many thanks guys and girls
June 15, 2011 at 3:29 pm
How did you manage to install SSRS on bothe Node 1 and Node 2?
Can you list the steps?
We have a clustered SQL Instance name – PRODRPT where the report server databases are hosted
We have installed RS on PRODRP01 (one node in the two node cluster)
When we try to install RS on PRODRP02 (second node in the two node cluster), we get the error in attached word document. Why can’t SSRS be pointed to the clustered SQL instance i.e. PRODRPT?
Am I missing something?
June 15, 2011 at 3:44 pm
This is an old thread, I would suggest starting a new one.
CEWII
June 15, 2011 at 3:53 pm
How did you install SSRS on the second node? and what are the steps involved?
We have a clustered SQL Instance name – PRODRPT where the report server databases are hosted
We have installed RS on PRODRP01 (one node in the two node cluster)
When we try to install RS on PRODRP02 (second node in the two node cluster), we get an error (in attached word document).
Why can’t RS be pointed to the clustered SQL instance i.e. PRODRPT?
Am I missing something?
Thanks
June 15, 2011 at 3:58 pm
Again, I would suggest starting a NEW thread, not trying to continue on with this one.
CEWII
Viewing 14 posts - 1 through 13 (of 13 total)
You must be logged in to reply to this topic. Login to reply