kubernetes storageat ebs speed,s3 costs

The performance and latency* of EBS with the storage economics of S3, exposed as an ordinary block device

Cost calculator

The same volumes and workload priced against AWS EBS gp3

5,000 GB
27 volumes

3 nodes × 9 volumes each · 10 × 484 GB cache disks per node

125 MB/s
3,000 IOPS

27,000 IOPS per node · stripe carries 30,000

Advanced

Capacity

80 %
30 % / month

1.5 TB rewritten / month per volume

10 % of capacity

500 GB per volume · 13,500 GB total

Throughputper volume

125 MB/s

5 TB written / month

125 MB/s

328.5 TB read / month

10 %
0.5 cores

7.5 cores per node

1 cores
4 GiB

31.3 GiB per node

10 disks

30,000 IOPS and 1,250 MB/s a node · 484 GB a disk, all of it on the free baseline

10 %

3 nodes at $1,275/mo — already the cheapest layout

Node architecture
Instance typesall 34

59%$10,750/mo saved

cheaper than EBS gp3, for 135 TB provisioned

backedblock.io$7,475/mo
EBS gp3$18,225/mo
Show calculation

backedblock.io

Compute
$1,275
Object storage$2,484
EBS cache$1,080
Cache performance$0.00
PUT requests$877
GET requests$355
Snapshots$1,404
Total$7,475

EBS gp3

Provisioned capacity$10,800
Throughput$0.00
IOPS$0.00
Snapshots$7,425
Total$18,225

Key features

  • Object storage economics, with EBS performance and latency*

    Capacity comes from S3 at S3 costs; the working set is served from a write-back cache on local EBS disks, and smart tiering keeps the hit rate high

  • Storage that speaks Kubernetes

    A CSI driver attaches volumes to pods over NVMe-oF/TCP, so provisioning, expansion, snapshots and restores are ordinary PVC operations

  • Snapshots to whichever S3 class is cheap enough

    Full and incremental, each an independently restorable point in time, sent to STANDARD, STANDARD_IA or GLACIER_IR per snapshot class

  • Granular control and QoS, per volume

    Read, write, and combined MB/s caps plus an IOPS ceiling — enforce performance limits exactly where you need them

  • Encryption and compression

    Encryption and compression keep data protected before it reaches S3, while reducing storage and network usage