Page 1 of 1
Configuring a computer to be a processing node for MPP (Wind
Posted: Fri Mar 20, 2009 6:33 am
by Soma123
Hi,
Do we need to follow this step of Installation Guide i.e. " Configuring a computer to be a processing node for MPP (Windows) " incase we are loading Datastage on a single CPU.
here we have installed Datastage 8 on Windows 2003 Server and now confused on should we go for this Configuration or not. We will have One WIn 2003 Server m/c loaded with IBM Information Server and other win XP machines (3-4) having Datastage Client Components accessing the Server machine. Does it resemble with SMP configuration ?
If anyone could advise on this that would be very much helpful.
Thanks in advance.
Posted: Fri Mar 20, 2009 1:49 pm
by DSguru2B
You configuration does not qualify for an MPP env. Server on one machine and x number of clients is a simple SMP env.
Posted: Fri Mar 20, 2009 4:28 pm
by ray.wurlod
I take it that this single-CPU server can deliver an infinite amount of processing, memory and disk I/O resources?
Posted: Sun Mar 22, 2009 11:35 pm
by Soma123
Well, this Win 2003 server m/c having 4gb of memory ais going to be used for training purpose of Paralle Jobs. I think a SMP configuration will be sufficient.
But the DataStage 8 installation guide does not talk about any such specific configuration for SMP? What should be done for configuring this server m/c for SMP?
Will the "Default" configuration file generated during installation is sufficient to carry out the task for SMP?
Otherwise incase I configure the computer to be a processing node for MPP, will that cause any harm?
Thanks.
Posted: Mon Mar 23, 2009 12:28 am
by ray.wurlod
An SMP configuration has the same fastname for all nodes mentioned in the configuration file. Provided that the default configuration file has two or more nodes it will be good for training in parallel jobs.
Posted: Mon Mar 23, 2009 6:34 am
by Soma123
Now I have created a configuration file having 2 Nodes to have SMP.
{
node "node1"
{
fastname "MACHINE1"
pools ""
resource disk "C:/IBM/InformationServer/Server/Datasets" {pools ""}
resource scratchdisk "C:/IBM/InformationServer/Server/Scratch" {pools ""}
}
node "node2"
{
fastname "MACHINE1"
pools ""
resource disk "C:/IBM/InformationServer/Server/Datasets" {pools ""}
resource scratchdisk "C:/IBM/InformationServer/Server/Scratch" {pools ""}
}
}
But why is it throwing so many Access Denied Error when I am validating it , like
C:/IBM/InformationServer/Server/PXEngine/etc/vsvars32.ksh 108: cannot open "C:/WINDOWS/TEMP/sh087C.aab" for output: Access is denied.
C:/IBM/InformationServer/Server/PXEngine/etc/vsvars32.ksh 108: C:/WINDOWS/TEMP/sh087C.aab: The system cannot find the file specified.
C:/IBM/InformationServer/Server/PXEngine/etc/vsvars32.ksh 131: cannot open "C:/WINDOWS/TEMP/sh087C.aad" for output: Access is denied.
thanks.
Posted: Mon Mar 23, 2009 2:55 pm
by ray.wurlod
Because the process doing the validation can not write to C:\Temp folder.
Posted: Tue Mar 24, 2009 4:27 am
by Soma123
Well after giving the permission now the Access denied error has gone. But now it's throwing a Warning "Virtual memory information is not available on this platform."
Now, where do I set up Virtual memory that will be visible to dataStage?