IT consultant David Linthicum says ignore claims on performance and just test your application in a target environment.
I’m always skeptical about claims that seem too good to be true, especially with regards to technology. The newest instances are claims made round the Google Cloud’s ability to give “performance stability,” as best highlighted in a commentary by Chandra Krintz, “Google Cloud’s Big Promise: Performance Stability.”
While cloud platforms, specifically, IaaS platforms, do their best to supply consistent performance, the performance that we see from most cloud providers is often a bit of “bursty.” They sometimes provide non-consistent performance inside the day. Here’s due largely to the undeniable fact that you finally share physical resources, comparable to CPUs, memory, disk, network, etc., with many alternative tenants, with a variety of requirements.
This is an old problem. Multiuser systems (versus multitenant) have long had an identical issues, and you’ll still see it today. As more users log in to a system, the slower the system becomes. Eventually, there’s a saturation point where the system begins to thrash, and function drops off significantly. Obviously, performance stability will be an outstanding thing.
[Need to know more about Google Cloud’s aspirations? See Google Compute Cloud Challenges Amazon.]
However, I’m more all in favour of how something is finished, beyond the assertions that it is usually done. Krintz states in her commentary:
Google Compute Engine is a better-generation of IaaS system and offers resource performance stability via a collection of novel engineering advances. These advances include: customized virtualization under KVM; advanced resource isolation technologies, corresponding to specialized Linux control groups that shield one process from others; clever data replication/ redundancy strategies; novel datacenter design and geographic placement; and dedicated high-speed fiber networks between well thought out and proven software services, reminiscent of App Engine, Cloud Storage, BigQuery, and YouTube.
The assertion is that Google, through its IaaS and PaaS services, provides consistent resource performance at very reasonable, when it comes to resources. Everything is healthier on this world, including redesigning applications to exploit the cloud-native features of the Google platform, in addition to the facility to give better and more consistent performance, specifically, better and more consistent than Amazon Web Services, that is the unstated objective of these claims.
The problem with this model is that, with a view to benefit from the consistent performance features, it’s really an application design, development, and deployment issue, greater than just good infrastructure. Thus, Google Compute Engine as an IaaS will never be as effective without the PaaS capabilities.
My suspicions? Will an unmodified application just dropped at the Google cloud platform be capable of make the most to the degree of benefit described? Google, as with other platforms, does require that its platform features be considered, to a point, when deploying applications in its cloud.
To be fair to Google, an identical issues exist with other IaaS or even PaaS players available in the market, including AWS, Rackspace, Microsoft, and others. Applications which can be modified to exploit native performance features of the cloud platforms they’re deployed upon will indeed offer up better performance and stability, including features comparable to auto-scaling. That’s why you hear the term “cloud native” a whole lot today. i think it’ll be a more popular approach as people who deploy to cloud-based platforms better understand the performance and stability benefits.
The claim that Google provides more consistent performance must be tested by each enterprise that moves applications and information to its cloud. i might recommend that they bring about a sequence of tests, using different loading profiles, reminiscent of simulating end-of-year processing using increasing loads on compute and storage, after which sustained loads for several days. This sort of testing will let you know lots.
As most folks who’ve migrated applications to the cloud know, the characteristics and profile of every application are more of a determination of performance than the platform. Thus, i’d also suggest that enterprises create test applications that leverage some cloud-native features, and observe the performance and stability differences over applications which are largely unmodified. On the grounds that you’re prone to move hundreds of applications and information sets to a cloud, this data may also help plenty in understanding the performance profile of your application, in addition to reducing the price of operations.
I doubt Google is the best provider that provides this benefit. However, Google’s deal with performance stability may get it more play available in the market over the following few years, because the second-place IaaS contender. Google is nice at making things fast, and that i suspect that its resulting IaaS and PaaS clouds won’t disappoint long-term. My experience have been that Google’s performance is sweet to great.
Keep in mind that, with any cloud-native features where applications are localized for specific cloud platforms, the tradeoff is portability for performance. In case you write cloud-native features on your application to support performance stability, that application should be modified when it’s moved to a different cloud platform, or even precipitated-premises. It is a cost and risk that should be considered before committing to a particular flavor of cloud computing.
Private clouds are moving rapidly from concept to production. But some fears about expertise and integration still linger. Also within the Private Clouds Step Up issue of InformationWeek: The general public cloud and the steam engine have more in common than you can think. (Free registration required.)
David S. Linthicum is senior vp of Cloud Technology Partners and an authority in complex distributed systems, including cloud computing, data integration, service oriented architecture (SOA), and massive data systems. He has written greater than 13 books on computing and … View Full Bio
More Insights