Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
suprasam
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
8 ms
·
1.
▲
by
suprasam
2mo ago
Thanks. The part that surprised me most was how well ZFS fits object storage and extended to pNFS too!
2.
▲
by
suprasam
2mo ago
This work was presented at LUG 2026 https://www.opensfs.org/wp-content/uploads/16-ZettaLane_LUG_...
3.
▲
Show HN: Open Lustre in the cloud, with ZFS OSTs on object storage not disks
(github.com)
1 points
by
suprasam
2mo ago
|
3 comments
4.
▲
ZFS, not FUSE: 13 GB/s from $0.02/GB GCS object storage on a single VM
(zettalane.com)
4 points
by
suprasam
7mo ago
|
0 comments
5.
▲
by
suprasam
9mo ago
ZeroFS doesn't exploit ZFS strengths with no native ZFS support, just an afterthought with NBD + SlateDB LSM Good for small burst workloads where everything kept it in memory for LSM batch writes. Once compaction hits all bets off with
6.
▲
by
suprasam
9mo ago
Yes that is the value prop. Cheap S3 instead of expensive EBS. EBS limitations: - Per-instance throughput caps - Pay for full provisioned capacity whether filled or not S3: - Pay only for what you store - No per-instance bandw
7.
▲
by
suprasam
9mo ago
For RDBMS pages on object storage - you might be thinking of Neon.tech. They built a custom page server for PostgreSQL that stores pages directly on S3.
8.
▲
by
suprasam
9mo ago
I hope you are not having that massive storage storage on public-cloud then you would need MayaNAS to reduce storage costs. For S3 as frontend use MinIO gateway - serves S3 API from your ZFS filesystem
9.
▲
by
suprasam
9mo ago
It is all part of ZFS architecture with two tiers: - Special vdev (SSD): All metadata + small blocks (configurable threshold, typically <128KB) - Object storage: Bulk data only If the workload is randomized 4K small data blocks - tha
10.
▲
by
suprasam
9mo ago
Yes, this is a core use case ZFS fits nicely. See slide 31 "Multi-Cloud Data Orchestration" in the talk. Not only backup but also DR site recovery. The workflow: 1. Server A (production): zpool on local NVMe/SSD/H
11.
▲
Native ZFS VDEV for Object Storage (OpenZFS Summit)
(zettalane.com)
127 points
by
suprasam
9mo ago
|
31 comments
12.
▲
by
suprasam
3y ago
New ZFS file system on object storage from https://www.zettalane.com
13.
▲
by
suprasam
3y ago
Experience very high-throughput > 3GB/s, even on a single VM instance with no traditional disk resources other than S3 object storage.
14.
▲
MayaNAS: ZFS on S3 object storage for high-throughput
(medium.com)
5 points
by
suprasam
3y ago
|
2 comments
15.
▲
by
suprasam
7y ago
Demo regarding unified ARC & Pagecache https://youtu.be/be0ph4b9vUE
16.
▲
Another ZFS Port on Linux
(crossmeta.io)
70 points
by
suprasam
7y ago
|
34 comments
17.
▲
by
suprasam
7y ago
The fun part is you can even develop from Linux using MinGW32 Cross Compile environment which produces native windows programs. Crossmeta FUSE also includes sshfs, fuse-nfs for remote file access and s3backer to connect to any S3 compatible
18.
▲
by
suprasam
7y ago
It has NO low level fuse api https://github.com/billziss-gh/winfsp/issues/94 Newer FUSE version 3 mostly address their limitations within Linux implementation. It adds support for concurrent writes and large
19.
▲
by
suprasam
7y ago
Crossmeta kernel has read/write support for XFS and EXT2/3/4 It has readonly support for reiserfs,Apple HFS+ All kernel drivers can be unloaded when not in use. From FUSE it has support for sshfs and NFS client V3 and V4.
20.
▲
by
suprasam
7y ago
The alternative project has no FUSE low-level API support. This is required to implement any real file system with FUSE on windows. Also they have dependency on cygwin project if your code requires some POSIX semantics, which they all do si
21.
▲
Announce Crossmeta FUSE for Windows
(crossmeta.io)
36 points
by
suprasam
7y ago
|
12 comments
22.
▲
MayaScale: Industry's First NVMe Fabric on Google Cloud for 10x EBS
(zettalane.com)
1 points
by
suprasam
8y ago
|
0 comments