Pillars of Premium Cloud Hosting
Security, Reliability, Speed, and Support. Multi-master, High-Availability, no Single Point of Failure, self-healing, Strong Cryptography, Global Reach, Highly Skilled 24×7 Support.
High-Availability
Strong Cryptography
Multi-master Clusters
Highest Performance
no Single Point of Failure
Business Critical 24x7 Support
High-Availability Network
Tier-1 Internet Service Providers
Exclusive collaboration with multiple Tier-1 ISPs.
N+1 Redundancy
At least N+1 Redundancy for all Network elements and duplicated links.
24x7 Monitor & Support
Our 24×7 Monitoring and NOC observes, work proactively to deliver a stable, congestion-free, premium network for our services.
High-Availability Power
Monthly Tests
Monthly tests are performed on all generators and UPS regarding building load, as well as load bank tests.
Transformers & Diesel Generators
At least N+1 pairs of transformers and at least N+1 diesel generators are installed, supporting at least twice or three times the total
Data Center load.
UPS Redundancy
At least N+1 and usually 2N UPS configurations are also in place in all Data Centers, with different levels of battery autonomy. The minimum autonomy is usually around 10 minutes.
High-Availability Hardware
No Downtime During Failure
If a hardware server in any of our clusters fails, then the services that were running on that hardware server get instantly transferred to other servers of the cluster.
Bypassing Time to Failure Observation
As a result, computing is not affected by the costly time it takes to observe failure and engage in time-consuming hardware replacement procedures.
No Downtime During Maintenance
No downtime is experienced during server maintenance since our clients’ resources and server configurations are temporarily transferred to separate nodes or clusters, depending on the type of maintenance.
Instant Hardware Upgrades
Since IT resources have already been made available during the cluster deployment, clients who wish to upgrade their resources can do it instantly.
High-Availability Storage
Triple Replicated Storage
For every GB used by our clients, we devote 2GB to replicate their data independently. In the unlikely event of double disk failure, the remaining master disk becomes Read-Only for the duration of the repair.
Self-Healing Storage
In case a disk fails so 1 out of 3 copies of data is lost, Storage will find available space on any other disk and will copy data from the other 2 replicas, re-enabling triple replication of the Storage. This is done automatically immediately after failures, ensuring maximum data reliability. During such events, the File System remains intact, so all the services that rely on the File System (like Data Bases) are not affected.
Pure SSD Storage
All servers’ critical data and software are served by pure, ultra-fast SSD storage, to avoid issues that can arise from reduced HDD reliability and speed. HDD storage is available once the server’s critical components are based on SSD Storage.
Ultimate Flexibility & Scalability in Cloud Computing
Hardware Failures Do Not Affect Server Uptime
Since MassiveGRID exclusively utilizes enterprise-grade IT equipment, it is highly unlikely for hardware failures to take place.
On top of that, our IT engineers make sure to monitor equipment performance and regularly proceed with maintenance, replacements, and upgrades.
In the unlikely event of hardware failure, clients’ servers instantly reboot on a healthy sector of the node or cluster, with identical resource capacity and configuration.
This way, we bypass all downtime that is caused by hardware failures, including the time needed to observe the failure, and manually set in motion a series of processes for the replacement of the defective hardware.
Horizontal & Vertical Cluster Expansion
Depending on the operational load some applications need horizontal scaling, while others vertical scaling. Our infrastructure architecture and technologies are designed to allow Horizontal & Vertical scaling with the following limits:
Vertical Scaling (per Cluster Node or Server):
up to 224 Physical, 448 Virtual Cores of Intel® Xeon® Platinum 8280L Processor at 4.00 GHz max turbo frequency
up to 128 TB of RAM per process
up to 4x 400 Gbps network
Horizontal Scaling:
up to 64 Nodes in a High-Availability Cluster
no limit on the number of H/A Clusters
Avoid Downtime By Maintenance
With IT downtime costing $300.000 per hour, on average, according to Gartner, it does not really matter If downtime is intentional or not.
Our Clusters offer flexibility since servers can move from one Node to another while live.
This means that if maintenance needs to be performed on your server, it can be easily moved to other nodes. As a result, maintenance is performed as the server is online. Upon maintenance completion, your server is instantly transferred back to the updated node.
Instant Upgrades Without Downtime
Upgrading available resources, such as Storage, CPU cores, RAM, or Network capacity is completed automatically.
Clients can deploy new High-Availability servers or upgrade specs of existing ones, instantly.
