Head of Product and Engineering at HTBase.
Love all things Datacenter and Cloud but love even more spending time with family and friends
Hey guys, we have recently announced JUKE, which gives you machine containers with native persistent storage and networking, including across clouds
It would be great to get your inputs and thoughts on the tool
Here is some info:
http://www.htbase.com/index.php/juke/
If interested, send me a message and Im happy to get you the download link
Thanks in advance
Bruno
Hey guys, we have recently announced JUKE, which gives you machine containers with native persistent storage and networking, including across clouds
It would be great to get your inputs and thoughts on the tool
Here is some info:
http://www.htbase.com/index.php/juke/
If interested, send me a message and Im happy to get you the download link
Thanks in advance
Bruno
@dafyre Its not free or open source right now. Its charged on a host/node base
Yep, it can be used independently from our hyperconverged solution
@dafyre HTFS from HTBase too! 
@Dashrender Yep, in that configuration we have 24 TB of SSD.
If we added an additional 32TB of HDD (given the fact you would already have 24TB of SSD), it would add around $13,000 to the total cost provided before
@Dashrender Processors are better in our case compared to the original thread.
To add 2 additional years of support, you would just be adding $5K, bringing the total to $76,600
Still a good price when compared
Would you mind if I add HTBase's HyperTask to the comparison?
Four node system with 2x Intel Xen E5-2650v4 per node, 256GB per node, 6TB of SSD per node.
Total cost (including 4 nodes and 3 years support):
$71,600
I think there are quite a few differences from an architecture perspective when you compare XenServer to KVM.
I agree that Xen still wins the performance battle depending on the workload and has more third-party tools available for management, specially for people who are learning that makes a difference.
But I do like the fact that KVM is part of the Kernel already and you can address hardware directly still, and get good performance with VirtIO, so its easier to make changes to it to adapt the hypervisor to what you want it to be or perform
With libvirt, latest I've seen this come more to a personal choice than anything
