Implementing SAP IQ at AWS

[former_member745605]

Hello All,

We are considering deploying an SAP IQ database to AWS using AWS Native Services.

The database would be less than 6 TB and be running SAP IQ 16.1.

Based on this note it is supported providing you meet OS and hardware sizing guide requirements.

2133194 - Can IQ run in a virtual or cloud environment? - SAP IQ 16.x https://launchpad.support.sap.com/#/notes/2133194

This appears to be reasonably straight forward for a simplex configuration. It seems a multiplex configuration might be better option to scale the database in the future but may be more complex to implement and support.

Are there any recommendations around using simplex vs multiplex configuration in AWS?

Is it supported to deploy a SAP IQ database in a multiplex configuration in AWS?

Has anyone deployed a production multiplex SAP IQ server to AWS using using AWS Native Services?

Also has anyone made use of the AWS EBS Snapshot feature for Disaster Recovery and database cloning?

Is there a recommended approach for a multi availability zone failover for an IQ database in AWS?

Is there a recommended approach for IQ database cloning in AWS.

Thanks,

BD

[Enda]

Welcome and thanks for visiting SAP Community to get answers to your questions. Check out our tutorial to get started in SAP Community: https://developers.sap.com/tutorials/community-start.html By adding a picture to your profile you encourage readers to respond to your question. Learn more about your profile here: https://developers.sap.com/tutorials/community-profile.html

[RolandKramer]

Hello B D
beside the comment of the "Rules of Engagement" check the Blog - SAP (Sybase) IQ – the hidden treasure … and simplex is fine for almost every scenario.

Deploying simplex/duplex on a Hyperscaler makes no difference to an on-premise implementation, which answers all of your questions ...

for the correct cfg file, check also the SAP Note 3017355 – SAP IQ 16.1 SP04 PLx – correct SAPIQDB.cfg settings

For additional Tipps and Hints see the SAP First Guidance – SAP NLS Solution with SAP IQ 16.x

Best Regards Roland

[markmumy · Accepted solution]

Hi BD,

For IQ (and ASE), we only certify to the OS. What that means, is that there is no technical certification needed for IQ to run on any platform so long as the OS is supported. For instance, if you are running SLES or RHEL Linux, IQ supports both of those.

The only piece that will be an issue is the Sybase IQ licenses. You should contact your SAP sales team to verify whether or not the licenses can be used on AWS. Generally, if the contract was with Sybase, you had to specifically purchase cloud deployment rights. If the contract is with SAP, then those rights are usually included. Only your sales team and make that call, though.

In terms of storage, there are many ways to define storage within AWS to meet the needs of a simplex or multiplex configuration.

Given that your database is just 6TB, you can use EBS or EFS storage. You would need to work with AWS as to which storage type is better given your other needs (snapshots, MZ availability, etc). The good news is that there is a CLI that can be used to create EBS snapshots. SO long as you have a CLI or script, then you can have IQ use the virtual backup methodology to perform the backup. an IQ virtual backup will call the script to perform the snapshot. It does it in a way that is safe for IQ and a way that we can guarantee data integrity of the backup and the subsequent restore.

Now, EBS may not be viable, directly, for a multiplex. I don't think an EBS volume can be mounted on two hosts simultaneously. That's where EFS comes in.

What has also come on to my radar is AWS Lustre Filesystem, or Amazon FSx for Lustre. This may have more of the features that you desire built in to the product. Something worth investigating. I am working with another customer and AWS to develop a solution for IQ Multiplex on AWS with Lustre. So far, I have not seen any roadblocks that would make this anything but a good solution.

For cloning databases, I would use the AWS utilities to snapshot the storage. The tools are quite good and efficient at moving around large amounts of storage. If not, then you would go with a more surgical approach and move objects, and data, from one IQ instance to another. Can be time consuming if you want most or all of the data.

Mark

[MarioDeFelipe]

Hello B D

Answering your question about native Amazon services, In case you want to deploy SAP IQ on AWS for the purpose of Data Tiering with NLS storage, the native solution from Amazon, S3 storage, might be used in combination with Data Intelligence

SAP Data Intelligence as cold store location, for SAP BW/4HANA Data Tiering Optimization (DTO) to store the cold data in Amazon S3. This only applies for BW/4HANA

You can check out more on this blogs;

https://aws.amazon.com/blogs/awsforsap/voice-of-customer-analytics-using-sap-data-intelligence-on-aw...

https://blogs.sap.com/2018/12/03/what-is-sap-hana-cold-data-tiering/

Also has anyone made use of the AWS EBS Snapshot feature for Disaster Recovery and database cloning?

EBS snapshots are expensive to be a recommended DR solution, if you calculate it for 6 TB, you will see a high cost, instead you might want to check and test CloudEndure Disaster Recovery for your SAP IQ, if you choose gp2/gp3 as the EBS storage, on the replication you don't need to select such a high performant storage, and use st1, see how it performs