mirror of
https://github.com/versity/scoutfs.git
synced 2026-07-19 22:42:40 +00:00
Auke Kok
72dc5695a6
This option adds a mount option, with default value of 16384, that adds an additional reserve amount of blocks for the meta device. The default value is 16384, which corresponds to 1GB of space, and just about doubles the internal value for the reserve that is calculated based on clients/mounts dynamically in sort of standard values. It also just compromises about less than 2% of the meta device size for the smallest meta device size. A suggested value for larger deployments is like somewhere around 256 blocks per GB of meta device size, i.e. 1/64 of the meta device space, and about 1.6% in effect. Customers who are running into issues can adjust their mount options to increase the value to have a larger safety buffer, or decrease it to potentially have a way to get out of low space conditions temporarily. Obviously one would want to increase the value of this option after resolving the low space condition issues as soon as possible. Our test suite will run with meta_reserve_blocks=0, so that the behavior of any of our tests is functionally unaffected by this change, and won't interfere with resolving underlying ENOSPC issues and their resolution. The addition of this option however allows us to artifically create ENOSPC conditions at will, and we may want to add tests specifically that do so. Signed-off-by: Auke Kok <auke.kok@versity.com>
Introduction
scoutfs is a clustered in-kernel Linux filesystem designed to support large archival systems. It features additional interfaces and metadata so that archive agents can perform their maintenance workflows without walking all the files in the namespace. Its cluster support lets deployments add nodes to satisfy archival tier bandwidth targets.
The design goal is to reach file populations in the trillions, with the archival bandwidth to match, while remaining operational and responsive.
Highlights of the design and implementation include:
- Fully consistent POSIX semantics between nodes
- Atomic transactions to maintain consistent persistent structures
- Integrated archival metadata replaces syncing to external databases
- Dynamic seperation of resources lets nodes write in parallel
- 64bit throughout; no limits on file or directory sizes or counts
- Open GPLv2 implementation
Community Mailing List
Please join us on the open scoutfs-devel@scoutfs.org mailing list hosted on Google Groups