Services Gravity vs. Data Gravity
Transcript
[Music] hey how's it going it's Keith Townsend from the CTL visor comm with another CTO DOS we're gonna talk about a topic that has been kind of floating in the cranium for the past year or so if you followed me on social media it's not a new concept to you but it may be data gravity is something that we've dealt with for the past several years ever since Dave McCrory formalized to thought that the more data that you have the harder it becomes to move the
compute or the data away from those centers of gravity you have to move your compute closer to your data it spawned I think a renaissance in the vdi in which you put your data in the centralized data center and then you move compute aka desktop images close to the data and that alleviate some of the challenges of users being able to get at the data that they need the most I want to introduce this concept of service gravity as cloud has taken hold you know we've
been concerned about moving our services or our datasets into the cloud and creating a data gravity in which we're locked into cloud providers because moving data out of the cloud moving data from cloud to cloud becomes cost prohibitive and sometimes technically impossible well as cloud computing begins to mature and companies adopt these services whether it's you know traditional storage services such as s3 which is our which is pretty easily replicated across the big cloud providers Azure and GCP offer s3 compatible storage you start to look
at services like GPU based machine learning and AI and which we can't recreate these services easily in our data center so we have two competing forces we have data gravity where the data may reside in our data center or in a cloud then we have this concept of service gravity as cloud providers become more more skilled at offering services that developers absolutely desire and that we can't recreate in our own data center at least not easily we end up getting locked into service providers not because
or not just because our data resides in there but because the services reside in there a great example is lamda if you build a functions base or server list based application and AWS it's not so simple to take that service and then move it to the servers like Azure or GCP and it's not because you know of the cold Python is Python whether that Python runs in land or the Python runs and as your cloud functions is is irrelevant it's the triggers the events that happen
off of the data and the services if you have s3 events or Cloud Gateway events that are tailored towards one service service portability becomes difficult so service gravity is now in this new form of IT lock-in that we have to worry about no longer are we just concerned about data and the amount of data stored in the cloud provider we're also concerned with services what services are we consuming from a cloud provider and can they be easily migrated to another service provider or even arm premises
so as you're figuring out your multi cloud strategy think about the services that you want to be portable those services that you want to build your applications around and whether or not you want to create or if the service gravity that you incur is worthwhile you want to learn more about the CTO adviser follow me on the web the CTO of Iser com or on Twitter at CTL visor talk to you next CTO DOS
