"automated provisioning tools like Jenkins or MAVEN."
Really? Would it kill you to spend 5 minutes with Google, looking up a few bits of jargon?
VMware CEO Pat Gelsinger last week introduced the company's second quarter results by saying the company has embarked on a “multi-year journey from a compute virtualization company to offer a broad portfolio of products driving efficiency and digital transformation.” And today at VMworld the company began to explain what that …
Getting a little tired of the constant harping on hyperconverged. Everyone seems to want to get in on this game, but its only really good for people who have a very steady software to compute ratio of systems, and on the whole this isn't a lot of businesses.
Some, if not most, database systems require TBs of storage all by themselves compared to the amount of CPUs and memory required. Lots of tiny web and app boxes require almost no storage whatsoever (but demand a lot of CPUs and memory).
I have never worked somewhere in my many decades career where the above was not the case.
Now you see companies like NetApp offering hyperconverged that isn't really hyper converged. Its just a bunch of VMware hosts with Storage hosts that look like hyper converged-sort-of-kind-of. It certainly sounds good to them.
NetApp's HCI product addresses the issues you raised, so it's probably not a good example of those HCI issues you mention. I'm not claiming it's better for everyone's use case, but with regard to HCI architecture issues you highlighted, it doesn't suffer from them.
It let's you unambiguously provision one or the other resource, without hidden or unpredictable resource consumption (and cost) where and when you don't expect it.
Disclosure: I work at NA
What I've settled on is pair of Xeons, max memory, some flash (NVMe, SSD, stick even) to get things rolling, with logging, and as you've said a filer. I've settled on BSD (for ZFS) for the latter. The only fly in the ointment is network which is only 1Gbps, something I'll be correcting here in the near future. Then things will be a bit more nimble.
Given that the flash is swappable, the hypervisor can be swapped as well. Hyper-V, VMware, bhyve, so far. All that memory is really handy for in-memory database work or plenty of VM's, or both. Analytics is a thang for me, as is machine learning/big data/predictive analytics/whatevah. Something I've been doing since '75. In this context, I'm still trying to figure out what HCI buys me. I've no problems seeing what VMware has on offer here. Completely out of my budget, still a really good place to be if I did have it.
I'll be waiting with bated breath to hear Trevor and the rest of the field people have to say.
Surely with a hyperconverged system you can allocate exactly what you need to what VM. So If it is low CPU, High storage then it fits great with a High CPU low storage VM.
As all the storage is tiered you can push the VM to large slow storage or high speed SSD or even mid-tier.
You can easily expand the storage, CPU or memory on the boxes and they also often give you high availability between all boxes so you don't need to worry about storage clusters and Server clusters and how they interact over metropolitan links. It can be automated so a box can failover and you can be sure that the storage and VMs will still be available.
Although it can create a a single box to go wrong, the HA feature would mitigate that and it is better than the multiple redundancies you'd need to put in place for a SAN - you lose you storage your server host is useless anyway. A HC box doesn't need hardware switches and cables to access the storage so there the box is less likely to get corrupt data with a network issue. It also does allow for a pick up and run type disaster scenario.
However you start getting enterprise scale and things would be difficult to run in HC for sure, but you can still run a reasonable operation on one.
"Azure will be king in 5-10 years. AWWho?"
Microsoft already went ahead of Amazon in annual cloud revenue run rate in the last quarter so they already are "King"
I think AWS will always be big in digital as long as it's primarily a Linux / Open Source focused world. But yes Microsoft are looking like the generic enterprise solution that will take the largest chunk of the market. However that digital focus could always change. According to Netcraft over 50% of websites run IIS and they have 10% of the top 1 million busiest sites. From personal experience I find IIS a hell of a lot easier to setup, load balance, distribute content and config on and manage than Linux / Apache / nginx etc. and I have experience of both. And IIS tends to be faster too.
I know for some reason people here don't seem to believe that Windows is often faster / more scalable than Linux, but don't take my word for it, see: https://www.rootusers.com/windows-iis-speed-test-benchmark-2017-results/
I suspect that Microsoft will eventually make all their base OSs free and charge for enabling advanced functionality and support. I think that's the best way for them to compete with Open Source.
"Microsoft already went ahead of Amazon in annual cloud revenue run rate in the last quarter so they already are 'King'"
The devil is in the details, as Microsoft's "cloud" revenue falls under it's "Intelligent Cloud" segment which sells significantly more product lines than just Azure services, many of which have nothing to do with XaaS or really anyone's definition of "cloud".
For reference, the Intelligent Cloud segment includes: "Server products and cloud services, including Microsoft SQL Server, Windows Server, Visual Studio, System Center, and related CALs, and Azure; (also) Enterprise Services, including Premier Support Services and Microsoft Consulting Services." https://www.microsoft.com/en-us/Investor/segment-information.aspx
Biting the hand that feeds IT © 1998–2020