Cloud
ScalarCloud Europe 2025 infrastructure: updates and improvements
Posted by brain in Cloud on 27 novembre 2024 with No Comments
Hello everyone, we have just validated the new ScalarCloud Europe 2025 infrastructures, and the hardware order was placed today.
The old ScalarCloud infrastructures (2021):
– Processor: 2 × Intel Xeon Gold 6242R – 20c/40t
– RAM: 768 GB ECC 2933 MHz
– Storage: 12 × 3.84 TB NVMe SSD (2 arrays of 6 NVMe SSDs in RAID 10)
Redundant connection: 2 x 10 Gb/s
Points of presence: UK (ERI), DE (LIM), FR (RBX, GRA)
The new ScalarCloud infrastructures (2025):
– Processor: 2 x Dual AMD Epyc 9354 – 64c/128t
– RAM: 1.5TB DDR5 ECC 4800 MHz
– Storage: 18 x 3.84 TB NVMe SSD (3 arrays of 6 NVMe SSDs in RAID 10)
Redundant connection: 2 x 25 Gb/s
Points of presence: FR (RBX, GRA, SBG), DE (LIM).
Health Certification HDS.
In summary: we will close the ERI (UK) site, replaced by Strasbourg, as Brexit has caused too many certification issues with some European clients.
The HDS (health certification) will become the default for our infrastructures.
We maintain the same hardware policy: no compromises on performance. The new AMD Epyc 9354 delivers better performance across the board and increases memory bandwidth from 140.8 to 460.8 GB/s.
On the network side, we move from a redundant connection standard of 10 to 25 Gb/s.
Finally, on the storage side, the standard of 2 RAID arrays is now upgraded to 3.
Currently, the 2 separate arrays allow clients who need ultra-performing and highly available local backup space to do so without relying on network services.
The 3rd array strengthens the setup with a spare array that allows more effective, secure, and faster switching in case of hardware failure:
1) Local migration to the spare array to immediately secure the data.
2) Migration of VMs to another node after data security.
3) Hardware maintenance on the initial node.
4) Testing and validation.
5) Returning to the initial node.
Not yet a client? Discover our managed Linux, PHP (5.6 to 8.3), MariaDB/MySQL, & Node.js solution starting at €145 excl. tax/month: StackX
Evolution of ScalarCloud Backup Infrastructure / PBS (Proxmox Backup Server)
Posted by brain in Cloud on 26 octobre 2024 with No Comments
Hello everyone, we are currently setting up a new backup infrastructure for our hosting clients (ScalarCloud).
To address the overlap issues that can sometimes occur, especially during nighttime network slowdowns on the OVH network currently in use, the new infrastructure will no longer be shared.
From now on, each node will have its own independent PBS (Proxmox Backup Server) as well as a dedicated 25 Gb/s interconnection for this purpose.
Have a great weekend, everyone!
Not a client yet? Discover our managed Linux, PHP (5.6 to 8.3), MariaDB / MySQL, & Node.js solution starting from €145 excl. VAT/month: StackX
Hypervisor Update
Posted by brain in Cloud on 23 décembre 2023 with No Comments
End-of-year update for all Proxmox VE hypervisors, which form the backbone of the ScalarCloud Public Cloud infrastructure.
Based on the tests we have conducted over the past month, we will for now retain the 7.4 branch of Proxmox VE for the Public Cloud’s production clusters.
We will schedule an update once all our tests are 100% conclusive. (In any case, there is no impact from your side for which this part is invisible).
No service interruption is expected, as machines are moved live from node to node during the updates before returning to their original location.
Increase in Local ScalarCloud SSD NVMe RAID Storage Capacities
Posted by brain in Cloud on 20 septembre 2023 with No Comments
Hello everyone, in response to the accelerated demand for space in recent months, we will soon be increasing the SSD NVMe storage capacity of our ScalarCloud platform.
These upgrades will be carried out in stages and node by node. No service interruptions are expected.
The process will be similar to hypervisor updates.
Your machines will be moved live and distributed to other standby nodes during the operation, and then return to their original location.
After that, a balancing via distribution will be performed: some virtual machines will retain their NVMe virtual storage on the first volume, and others will be moved (live) to the second volume.
These operations will start on October 3rd and aim to double the current RAID 10 storage capacity by adding a new matrix, allowing us to increase from 24 TB NVMe per node to 48 TB.
The end date of these operations is not yet known, as it requires coordination with OVHcloud node by node for the hardware upgrades. We will keep you informed when these operations are completed.
Have a great end to your weekend.
Recent Comments