November 2019: The advent of the HC3 Edge Fabric
A distributed bridge. Relaxing of the strict network topologies for the backplane network, meaning no backplane switches needed for the HC3 cluster, especially wanted at the EDGE (Retail), either using Ethernet or 15Gbit Thunderbolt direct connect between the nodes in a ring (inherently redundant), freeing up all the ethernet ports for public traffic. Brilliant!
November 2019: HC3 Rest API, Ansible playbooks, Multi-cluster-monitoring explained
Distributed computing consisting of hundreds or thousands of clusters requires remote lifecycle management of the distributed virtual payload, in the case with HC3 orchestrated using Ansible and home grown multi-cluster monitoring technology.
May 2018: Cinder and Scale HC3 for environments with high performance demands
NVMe is designed for unparalleled IO performance and Scale Computing’s new HyperCore-Direct NVMe integrated architecture is designed for environments that have intense data operations, high performance applications, and other exceptional performance requirements. We have now integrated HyperCore-Direct directly with Cinder to deliver that performance to the Openstack community. Phil White, CTO at Scale Computing brilliant as always.
September 2017: A whiff of NVMe future potential with Scale HC3 Direct, and GCE based HC3 Cloud Unity
Phil White, CTO at Scale Computing, reviews how their SCRIBE storage engine works for their HCI solution. This was designed for scalability and efficiency from the start, with no file system, storage protocols, or VMs in the data path. SCRIBE has been integrated directly to the hypervisor, which has provided near bare metal performance with the adoption of NVMe. Phil introduces HyperCore, their NVMe over fabrics HCI solution based on the HC3 software stack.
At 26:00 in the video above Phil interestingly mentions ”The Attack Of The Microseconds” -document, found here, the movie found below:
Phil White, CTO at Scale Computing, announces details of Cloud Unity with Google Cloud. This effectively gives seamless DR with full failover in the cloud for anything already running on their HC3 platform.
April 2016: The HEAT is on, Dynamic Tiering and IOPS display GA in Version 7.
It is time to turn up the HEAT and slide into something more comfortable: Scale HC3 adaptive dynamic tiering is finally here. Enjoy tiers without fears, tiers without tears. This is not simply a cache in flash, it is a per VM disk dynamically assignable location designation disk block priority based on per block IOPS activity. Once again the brilliant development team at Scale has made something unbearable complex into something simple comfortable and useful. This is what it looks like:
June 2015: SCRIBE goes FLASH and gets dynamic tiering.
Upptagningen nedan är på engelska och innehåller en brilliant utläggning av Scale’s Chief Architect Phil White och Scott Loughmiller VP of Engineering där de redogör för nyheter i SCRIBE. Inspelningen gjordes den 25 Juni 2015 på Virtualization Field Day 5.
- Scott at 04:03: A lot of the stuff we talk about is going to sound familiar with Scale, it is going to sound like things you already know of from other products. But there is something a little bit unique, and little bit special in each one of these parts, and when you put them together in concert it makes the whole thing really special. So when you you have that feeling, that this is just like what you already know, ask why it is different, because it is.
- Phil at 25:23 ”This year we release FLASH friendly VSD’s” A write-back-cache and log based journal for the Virtual Storage Devices (VSD) doing properly aligned IO with ability to coalesce small blocks into large blocks hence making the VSD’s SSD friendly. Thanks to the modularity in SCRIBE the journals are wrapped around already existing VSD’s in the runnnig Scale Cluster. The modular design also offer the ability to having the journal removed at some point in the future if needed. Future RSD placement algorithms such as erasure coding will work seamlessly with these VSD journals as well as the need for improving write-performance of those arise.
Under Virtualization Field Day 5 den 25 Juni 2015 hölls även denna uppdaterade introduktion till Scale HC3 där Scott Loughmiller introducerar oss till Scale och vad som gör HC3 unikt på virtualiseringsmarkanden (tip: SIMPLICITY on every level):
January 2015: The year of a new face and cross cluster DR.
Upptagningen nedan är på engelska och innehåller en presentation av Scale’s Chief Architect Phil White. Inspelningen gjordes den 16 January 2015.
1) Scale Computing DR (Per VM, Disaster Recovery) djupdykning.
Nästa video visar Scale GUI och Replikering och Rollback av VM mellan två kluster:
Contact Data Resilience AB if you would like to experience a live demo of the Scale Computing platform.
2014: The year of SCRIBE and Rolling Upgrades.
Här nedan finner ni fyra videoinspelningar som beskriver Scale HC3. Upptagningarna nedan är på engelska och presenteras av Scale’s CTO Jason Collier (1), VP of Marketing, Dave Demlow (2) och Chief Architect Phil White (3&4). Inspelningen är från den 23 April 2014 i San Mateo CA.
1) Vilka är Scale Computing?
2) Vad är HC3?
3) Hur fungerar Scale HC3?
4) Hur fungerar Scale HC3’s lagring och SCRIBE?