Virtualization Technology News and Information
VMblog Expert Interview: Applications, Availability and the Right Tool for the Job

Interview SIOS Meltzer 

VMblog recently had the chance to catch up with Jonathan Meltzer, Director, Product Management, at SIOS Technology to talk about high availability and using the right tools for the job.

VMblog:  Tell us what you mean when you say companies should consider deploying a selection of tools to respond to company- and application-specific availability challenges.

Jonathan Meltzer:  When your application fails, you don't necessarily have to use a full cluster failover to restore operations. Failover is the biggest hammer in your toolbox. However, advanced clustering solutions offer several, more precise actions, or HA tools, that can be implemented before using the big hammer of failover. These actions, which may include restarting services or rebooting the server, can often quickly restore application availability.  

VMblog:  When does it make sense to use tools other than a hammer to address availability challenges?

Meltzer:  There is a famous saying that says that if all you have is a hammer, then everything looks like a nail. For critical applications, you want to invest in a clustering software with the ability to provide that "hammer" of full failover when you need it. But it should also be able to restart services and perform other more precise steps before going directly to the failover step. 

In other cases, the criticality of the application doesn't justify the cost of the high availability "hammer". This is where other tools outside the "hammer" may be appropriate to address application availability at a lower cost. 

VMblog:  What are some examples of availability tools fit for less mission critical applications and how do they work?

Meltzer:  Sophisticated failover clustering software can use a sequence of steps in addressing a failed application. It can start by restarting application services to allow faster restoration of services and lets you save full failover for when it's needed.

If your application isn't critical enough to ever justify the need for a failover, you can also apply availability software that only restarts services, and if that doesn't work, restart the entire server or cloud instance.

VMblog:  For what applications are ‘hammers' needed to ensure high availability?

Meltzer:  Typically, we see companies with mission-critical applications like SQL Server, SAP S4/HANA and Oracle consistently needing high availability. However, note that while they need the failover "hammer" sometimes, they should be protected with clustering software that includes the more precise tools that enable it to restart services first, and then failover only when necessary. 

VMblog:  What requirements should ‘hammers' meet to prevent a disaster from application failure, and how do they work?

Meltzer:  A failover clustering software should be application-aware. That is, it should have hooks into specific applications that monitor for faults and orchestrate failover in compliance with best practices for that specific application. For example, SAP contains multiple parts that need to be located on specific servers and started up in a prescribed order. Clustering software should automatically set up those conditions as part of configuration - without manual scripting - and maintain those best practices when it fails over. So, in a manner of speaking, even the "hammer" in a good clustering software will failover with precision.

VMblog:  What are the biggest benefits to using the right tool at the right time?

Meltzer:  By applying the right availability and high availability tools enables IT to achieve the optimum balance between cost and required recovery time.

VMblog:  Should customers select all of its high availability tools from the same vendor, or can they mix and match?

Meltzer:  While it would be ideal to select all availability tools from the same vendor, every implementation is different, so sometimes yes and sometimes no. Customers should think about the criticality of their application - RTO (Recovery Time Objective), or how much it costs the business for every minute it takes to recover the application and RPO (Recovery Point Objective), or how much data you can afford to lose. They should also think about the orchestration of the application failover step separately from the replication of their data and choose the right combination of application protection and replication to meet their needs.  

VMblog:  What High Availability tools does SIOS have in its toolbox to offer customers?

Meltzer:  SIOS Protection Suite includes application-aware failover clustering software called SIOS LifeKeeper that restarts services first (precision tools) and fails over (hammer) only when necessary. It monitors the entire stack and application and ensures failovers happen in complete compliance with best practices for the application - even if the application is a complex, multi-part database or ERP system like SAP HANA or SQL Server. The SIOS Protection Suite also includes efficient block-level replication software called DataKeeper that can be used with LifeKeeper or on its own to synchronize local storage in failover clustering environments to create a so-called SANless cluster. SANless clusters eliminate the shared storage requirements of traditional clustering software, enabling failover clustering in public, private, and hybrid cloud environments.


Published Tuesday, July 07, 2020 7:32 AM by David Marshall
Filed under: ,
VMblog Expert Interview: Applications, Availability and the Right Tool for the Job – COVID-19 Job Impact - (Author's Link) - July 29, 2020 9:19 PM
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<July 2020>