e012fca9a2 mount: keep lookups from stalling behind directory cache rebuilds (#12754)
* mount: release the meta cache lock between batches of a rebuild wipe

deleteFolderChildrenForRebuild held mc's write lock while it listed every
cached child of the directory and deleted them. Every metadata read in the
mount (FindEntry, IsNameFresh, isOversized) takes the read lock, and a
waiting writer blocks new readers, so while a large directory was wiped,
lookups and getattrs in unrelated directories queued behind the scan. The
scan gets slower as leveldb accumulates the deletion markers of earlier
wipes of the same directory, which happens quickly once lookups rebuild
expired directories.

Wipe rebuildWipeBatchSize (1024) children per lock hold and drop the lock
between batches. A child is judged pinned under the same lock that deletes
it, so a local-only create is kept whether it was pinned before the wipe
started or between batches. The listing that refills the directory already
ran outside the lock, so batching adds no interleaving the build did not
allow.

Add LevelDBStore.BatchDeleteEntries, the counterpart of BatchInsertEntries.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

* mount: bound how long a lookup waits on the directory rebuild it starts

Since #11268 a lookup whose parent's cache expired rebuilds the parent and
every expired ancestor up to the mount root before it answers, and every
lookup beneath them joins the same builds. With a slow listing (a large
directory, a busy filer, or the meta cache lock held by another rebuild) a
plain stat, an ls -l, or the parent touch after mkdir waited for all of it.

Wait at most lookupRebuildTimeout (2s), then fall back to one LookupEntry
RPC. Since #12206 the shared build does not run on the requester's context,
so it keeps going after the waiter gives up and later lookups find the
directory cached. A timed-out wait records a rebuild attempt, so lookups
during the cooldown do not queue on a build that is already known to be
slow.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

* mount: wait for the rebuild's listing in TestLookupDoesNotWaitOutSlowRebuild

The build runs on its own goroutine, so the lookup that stopped waiting
for it can return before the build's ListEntries reaches the filer. The
test read `started` without waiting right after the lookup returned, which
relied on an ordering nothing guarantees, and the ListEntries count taken
afterwards assumed the same. Wait for the listing to arrive, bounded by
the test's deadline.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

* mount: keep children inserted while a rebuild wipe runs

A deferred create inserts its placeholder before Create registers the
handle that pins it. The single-lock wipe made such an insert wait until
it finished, so the placeholder survived; a batch that runs in that gap
judged it unpinned and dropped it, and the rebuilt listing then omitted
the file until its flush.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* mount: drop a cached path that an event vacates during a rebuild wipe

A build replays only the events later than its listing's snapshot and
leaves the earlier ones to the listing, which starts after the wipe. With
the wipe in batches a file's whole life fits inside it: created, flushed,
then unlinked or renamed away. Both events are earlier than the snapshot
and are skipped, the listing has no such file, and the placeholder the
wipe kept stays in the published directory until the next rebuild.

While a directory's wipe runs, an event buffered for it now takes the
path it vacates out of the store at once, fenced by versions like any
other event. After the wipe the store holds what the listing put there,
and a removal earlier than the snapshot may still arrive, so from then on
the event is left to the replay as before.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

* mount: check which file a removal took away during a rebuild wipe

Two things the removal applied during a rebuild wipe got wrong.

A placeholder is unversioned, so the version gate lets every removal
through. One that arrived late, for an earlier file of the same name,
took out the placeholder of a file created since. The entry the event
took away is now compared with the cached one, by inode or, with no
inode to compare, by creation time. An event that only names the file
cannot tell and is taken at its word.

The entry was deleted without OP=MV, so taking one name of a hard-linked
file out of the cache also took one off the count its other names share.
A replayed event that vacates a path already deletes with OP=MV.

Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>

* mount: keep a removal the rebuild wipe declined out of the replay

The wipe declines a removal when what is cached there is newer than the
event or is another file than the one it took away. The event stayed
buffered anyway. An unversioned listing clears the children's version
records and replays every buffered event, so the declined removal then
deleted the entry the wipe kept and the directory was published without
it.

A declined removal now buffers only the name it puts in place, so a
rename still lands on replay.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* mount: check a declined removal again when the build replays it

The previous commit dropped the removal half of an event the wipe
declined, judged by the file cached when the event arrived. Earlier
buffered events can replace that file during the replay: a listing
attempt that failed leaves its events buffered, so the retry's wipe may
see the first attempt's file while the replay puts a later one in place
first. The dropped delete then left that later file published.

The wipe now records which removals it declined, and the replay checks
each against the entry it finds at that point, replaying only the name
the event puts in place when that entry is still another file. A
removal declined by version is replayed as before.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* mount: keep a declined removal's invalidation off the placeholder's handle

After the replay the build invalidates every buffered event again,
raised to the snapshot, and it still sent the original of a removal the
replay rejected. A late delete of an earlier file then reached the
handle of the placeholder at that name: the placeholder is unversioned,
so the raised timestamp got past it and the handle was marked deleted.

The replay's decision on a declined removal now replaces the event in
the buffer, and the invalidations use what was replayed. It is made
before the snapshot check, so a removal the listing covered is judged
the same way.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* mount: keep an event's signatures on the halves a rename is split into

A declined removal replays as its rename's create half, and a rename across
a building directory is split in two. Both dropped the event's filer
signatures, so a handle stamped by that filer no longer fenced the replayed
invalidation by timestamp and could take older listed contents over newer
ones it had read.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* mount: drain invalidations around the declined-removal build check

Invalidations run on a background worker, so the baseline could miss the
removal's own and the final read could run before the build's arrived.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: MapleWhiteZZZ <18653045+MapleWhiteZZZ@users.noreply.github.com>
Co-authored-by: Claude Fable 5.2 <noreply@anthropic.com>
Co-authored-by: Chris Lu <chrislusf@users.noreply.github.com>
2026-11-19 00:04:03 -08:00
2026-11-19 02:58:55 +00:00
2019-04-30 03:23:20 +00:00
2023-01-05 11:01:22 -08:00

SeaweedFS

Slack Twitter Build Status GoDoc Wiki Docker Pulls SeaweedFS on Maven Central Artifact Hub

SeaweedFS Logo

SeaweedFS is a simple and highly scalable distributed file system. There are two objectives:

  1. to store billions of files!
  2. to serve the files fast!

One SeaweedFS cluster serves an S3 object store, a POSIX file system, a lakehouse with S3 Tables and a built-in Iceberg REST Catalog, and vector search with S3 Vector Buckets, all on shared storage. Each blob is one disk read away, capacity grows by starting another volume server, and cloud storage can be cached or tiered transparently. Blob reads and writes have O(1) complexity and can run at the full speed supported by the underlying hardware.

SeaweedFS feature overview: S3 objects, POSIX files, table buckets with an Iceberg REST Catalog and Lance namespace, and S3 Vector Buckets share one cluster of masters, filers, and volume servers, with cloud caching, tiering, and cross-cluster replication.

Table of Contents

Quick Start

One command

Download the latest binary from the releases page and unzip the single weed (or weed.exe) file, or let the install script put it in /usr/local/bin:

curl -fsSL https://raw.githubusercontent.com/seaweedfs/seaweedfs/master/install.sh | bash

For SeaweedFS Enterprise binaries, pass --enterprise to the script: ... | bash -s -- --enterprise.

Then start a ready-to-use S3 object store:

AWS_ACCESS_KEY_ID=admin \
AWS_SECRET_ACCESS_KEY=secret \
S3_BUCKET=my-bucket \
./weed mini -dir=./data

That's it. The S3 endpoint is at http://localhost:8333, my-bucket exists, and admin/secret are valid credentials:

AWS_ACCESS_KEY_ID=admin AWS_SECRET_ACCESS_KEY=secret \
  aws --endpoint-url http://localhost:8333 s3 cp README.md s3://my-bucket/

The same process also runs the master, a volume server, the filer, WebDAV, the Iceberg REST catalog, and the Admin UI. Add S3_TABLE_BUCKET=warehouse to also create an Iceberg table bucket, or warehouse:LANCE for a Lance one. Drop the AWS keys to run without authentication for development.

Without Admin authentication or mTLS, weed mini binds the Admin UI/API and its worker gRPC control plane to loopback rather than -ip.bind. Set WEED_ADMIN_PASSWORD or configure https.admin mTLS to keep the network bind. Remote workers must opt in with -admin.worker.ip=<address> and should configure grpc.admin mTLS. -admin.allowInsecureBind restores the legacy unauthenticated network bind and should only be used on an isolated network.

macOS: if the binary is quarantined, run xattr -d com.apple.quarantine ./weed first.

weed mini is auto-tuned for one node and is fine for single-node production, such as an S3 gateway that issues presigned URLs. See Quick Start with weed mini.

Docker

docker run -p 8333:8333 -v weed-data:/data \
  -e AWS_ACCESS_KEY_ID=admin \
  -e AWS_SECRET_ACCESS_KEY=secret \
  -e S3_BUCKET=my-bucket \
  chrislusf/seaweedfs

Same behavior as the weed mini command above.

Docker Compose

To run master, volume server, filer, S3, and WebDAV as separate services:

wget https://raw.githubusercontent.com/seaweedfs/seaweedfs/master/docker/seaweedfs-compose.yml
wget -P prometheus https://raw.githubusercontent.com/seaweedfs/seaweedfs/master/docker/prometheus/prometheus.yml
docker compose -f seaweedfs-compose.yml -p seaweedfs up

Docker Compose for S3 adds credentials, and the docker/compose folder has variants for replication, mounts, message queues, and more.

Kubernetes with Helm

helm repo add seaweedfs https://seaweedfs.github.io/seaweedfs/helm
helm install seaweedfs seaweedfs/seaweedfs -n seaweedfs --create-namespace -f values.yaml

A production-shaped values.yaml for a three-node cluster: two copies of every write, three masters, and an S3 endpoint with credentials and a bucket.

global:
  seaweedfs:
    enableReplication: true
    replicationPlacement: "001"   # one extra copy on another server; "002" for two

master:
  replicas: 3
  data:
    type: persistentVolumeClaim   # the cluster's default storage class; add storageClass to pick one
    size: 1Gi

volume:
  replicas: 3                     # at least 1 + the sum of the replication digits
  dataDirs:
    - name: data
      type: persistentVolumeClaim
      size: 500Gi
      maxVolumes: 0               # size the volume count from the disk

filer:
  replicas: 2
  data:
    type: persistentVolumeClaim
    size: 20Gi

s3:
  enabled: true
  replicas: 2
  enableAuth: true
  credentials:
    admin:
      accessKey: admin
      secretKey: change-me
  createBuckets:
    - name: app-storage

The S3 endpoint is the seaweedfs-s3 service on port 8333. Helm Chart Recipes has values for a development cluster, a lakehouse with the Iceberg catalog exposed, filer metadata on PostgreSQL, and node-local disks. The SeaweedFS Operator and the CSI driver are the other Kubernetes paths.

Build from source

git clone https://github.com/seaweedfs/seaweedfs.git
cd seaweedfs/weed && make install

weed lands in $GOPATH/bin. Getting Started covers running master, volume, filer, and S3 as separate processes.

Scale out

Capacity is a volume server. Start one on any machine with disk and point it at the master:

weed volume -dir=/data -master=<master_host>:9333

Nothing rebalances until you ask it to. Throughput is a filer or S3 gateway; they are stateless, so run as many as you need behind a load balancer. Production Setup walks through a multi-node cluster.

Back to TOC

Why SeaweedFS

Fast

  • One disk read per blob. A small file is one blob; a large file is split into chunks of a few MB, each its own blob. A volume server keeps a 16-byte index entry per blob in memory and reads it in a single seek, also for erasure-coded data.
  • The master is not in the read path. Clients cache the volume-to-server mapping and talk to volume servers directly.
  • 40 bytes of metadata per file on disk. Small files are packed into append-only volume files, so there is no per-file inode, no per-file metadata file, no fragmentation, and writes are SSD friendly.
  • Hot data is replicated; erasure coding is applied to warm data in the background, so writes never pay the encoding cost.
  • The Rust volume server is a drop-in for higher throughput and lower tail latency on the same on-disk format.

On one laptop, weed benchmark writes 1KB files at 15,700 per second and reads them back at 47,000 per second, and a mixed S3 warp run totals 3.2 GiB/s. Numbers are in the Benchmark section; throughput grows with volume servers and gateways.

Scalable

  • The master tracks volumes, not files. A cluster with billions of files has a few thousand volumes, so the master stays small. One master is enough for most clusters; run three for Raft failover.
  • Adding a server adds capacity with no data reshuffle. Balancing, vacuum, erasure coding, and repair run on demand from weed shell or the maintenance worker.
  • Filer and S3 gateways are stateless and scale linearly. Directory metadata lives in a store you already run: LevelDB, RocksDB, SQLite, MySQL, PostgreSQL, Cassandra, HBase, MongoDB, Redis, Elasticsearch, etcd, TiKV, FoundationDB, YDB, ArangoDB, Tarantool, and MySQL or PostgreSQL compatible databases such as TiDB, CockroachDB, and MemSQL.
  • Rack and data center aware replication, tiered storage across disk types, and transparent cloud tiering for unlimited capacity.
  • Files from a byte to tens of TB. Volumes up to 8TB with the large-disk build.

The most complete S3 API

The S3 gateway implements the object, bucket, S3 Tables, S3 Vectors, IAM, and STS APIs on one endpoint, so the AWS SDKs and CLI, rclone, restic, Spark, and Trino work unchanged. S3 Vectors uses the weed-vector service described below.

API Operations
S3 bucket and object 73
S3 Tables 36
S3 Vectors 19
IAM 39
STS 5

The full operation list is in Amazon S3 API, and Supported APIs vs MinIO compares. The S3 compatibility suite and the SDK, IAM, SSE, policy, and Spark integration tests run in CI on every change.

A data warehouse with S3 Tables

SeaweedFS is a lakehouse in one system. S3 Table Buckets hold Apache Iceberg tables by default, or Lance tables for vectors and multimodal data, and the built-in Iceberg REST Catalog and Lance namespace serve them directly. The S3 Tables API and Iceberg REST Catalog share the same namespaces, table metadata, and permissions. There is no Hive Metastore, Glue, or separate catalog service to deploy, secure, and back up.

  • Query engines operate on the same tables at the same time: Spark, Trino, Dremio, DuckDB, Apache Doris, RisingWave, ClickHouse, and LanceDB. Catalog commits are atomic compare-and-swap, so concurrent writers are safe. Lakekeeper can front the same storage with STS-vended credentials. OLake Go is a high-performance, open-source EL (Extract–Load) platform that bridges operational databases and open lakehouse storage - writing Apache Iceberg tables or Parquet files on Amazon S3.
  • Automated table maintenance: compaction, snapshot expiration, orphan file removal, and manifest rewriting, configured per bucket or table through the S3 Tables maintenance APIs, and the same for Lance.
  • IAM at the bucket, namespace, and table level with standard bucket policies, see S3 Tables Security.
  • A Hadoop compatible file system for Spark, Flink, and HBase.

S3_TABLE_BUCKET=warehouse ./weed mini -dir=./data brings the whole stack up on a laptop.

Vector search with S3 Vector Buckets

Keep embeddings alongside your objects and tables in the same SeaweedFS cluster. S3 Vector Buckets expose the AWS S3 Vectors API for semantic search, retrieval-augmented generation (RAG), and recommendations, using Lance tables on SeaweedFS storage.

  • Create vector buckets and indexes; insert, retrieve, list, and delete vectors with the AWS SDKs or CLI pointed at your S3 endpoint.
  • Find the nearest vectors with cosine or Euclidean distance, filter by metadata, and return distances and metadata with the results.
  • Control access with s3vectors:* IAM actions and vector bucket policies; tag buckets and indexes for organization and access control.

Install the separate vector engine; the default installer installs only weed:

curl -fsSL https://raw.githubusercontent.com/seaweedfs/seaweedfs/master/install.sh | bash -s -- --component vector

Release 4.53 provides vector binaries for Linux amd64 and arm64. If your release has no binary for your platform, build from source. From a directory where you want a new checkout:

git clone https://github.com/seaweedfs/seaweedfs.git
cd seaweedfs
(cd seaweed-vector && cargo build --release -p weed-vector-server)

If you already have a checkout, start at its root and skip the clone and cd seaweedfs steps.

For Windows source builds, set PROTOC to a native protoc.exe before running Cargo; the desktop build workflow shows how to select the vendored executable.

Run the installed weed-vector binary alongside the local cluster started with weed mini above (or use seaweed-vector/target/release/weed-vector after building from source):

weed-vector --masters=localhost:9333

It registers with the master and the S3 gateway discovers it automatically. An AWS S3 Vectors client uses the same S3 endpoint at http://localhost:8333. The vector engine stores its data through the cluster's filers and volume servers, so embeddings share the storage layer's replication and erasure coding. See the AWS SDK examples for bucket, index, and query operations.

A fast cache for cloud storage

Cloud Drive mounts a bucket from S3, Google Cloud Storage, Azure, Backblaze B2, Wasabi, Storj, or any S3-compatible store into SeaweedFS and serves it at local speed:

  • Metadata is pulled once, so listing, stat, and directory walks cost no cloud API calls.
  • File content is downloaded once, on first read or warmed by folder, name pattern, size, or age, and cached with the capacity of the whole cluster: cache everything, no churn.
  • Local writes complete at local latency and are written back to the cloud asynchronously in the cloud's native layout, so other tools keep reading the bucket directly.
  • Uncache by the same rules to free local disk while keeping the metadata.

Cloud Tier goes the other direction, moving whole warm volumes to cloud storage while keeping one-read access, and the Gateway to Remote Object Storage mirrors every bucket to a remote store. Faster and cheaper than reading the cloud directly.

Active-active replication and more

Back to TOC

Architecture

SeaweedFS Architecture

  • Master servers, one or a Raft group of three, track which volume lives on which volume server and hand out file ids. They are not in the read path.
  • Volume servers store blobs in append-only volume files, keep a 16-byte in-memory index per blob, and replicate or erasure-code at the volume level.
  • Filer servers add directories and files on top, with metadata in a store of your choice, and expose HTTP, S3, WebDAV, SFTP, FUSE, and the table catalogs.

The blob store started from Facebook's Haystack, erasure coding takes ideas from f4, and the whole has a lot in common with Tectonic and Colossus. How file ids are assigned, written, and looked up, and why a master that tracks volumes scales, is in Blob Store Architecture; the services are in Components and the white paper.

Back to TOC

Compared to Other Systems

Most other distributed file systems seem more complicated than necessary.

SeaweedFS is meant to be fast and simple, in both setup and operation. If you do not understand how it works when you reach here, we've failed! Please raise an issue with any questions or update this file with clarifications.

SeaweedFS is constantly moving forward. Same with other systems. These comparisons can be outdated quickly. Please help to keep them updated.

Compared to HDFS

HDFS uses the chunk approach for each file, and is ideal for storing large files.

SeaweedFS is ideal for serving relatively smaller files quickly and concurrently.

SeaweedFS can also store extra large files by splitting them into manageable data chunks, and store the file ids of the data chunks into a meta chunk. This is managed by "weed upload/download" tool, and the weed master or volume servers are agnostic about it.

Compared to GlusterFS, Ceph

The architectures are mostly the same. SeaweedFS aims to store and read files fast, with a simple and flat architecture. The main differences are

  • SeaweedFS optimizes for small files, ensuring O(1) disk seek operation, and can also handle large files.
  • SeaweedFS statically assigns a volume id for a file. Locating file content becomes just a lookup of the volume id, which can be easily cached.
  • SeaweedFS Filer metadata store can be any well-known and proven data store, e.g., Redis, Cassandra, HBase, Mongodb, Elastic Search, MySql, Postgres, Sqlite, MemSql, TiDB, CockroachDB, Etcd, YDB etc, and is easy to customize.
  • SeaweedFS Volume server also communicates directly with clients via HTTP, supporting range queries, direct uploads, etc.
System File Metadata File Content Read POSIX REST API Optimized for large number of small files
SeaweedFS lookup volume id, cacheable O(1) disk seek Yes Yes
SeaweedFS Filer Linearly Scalable, Customizable O(1) disk seek FUSE Yes Yes
GlusterFS hashing FUSE, NFS
Ceph hashing + rules FUSE Yes
MooseFS in memory FUSE No
MinIO separate meta file per drive for each file Yes No
RustFS separate meta file per drive for each file Yes No

GlusterFS stores files, both directories and content, in configurable volumes called "bricks". It hashes the path and filename into ids, and assigned to virtual volumes, and then mapped to "bricks".

Compared to MooseFS

MooseFS chooses to neglect small file issue. From moosefs 3.0 manual, "even a small file will occupy 64KiB plus additionally 4KiB of checksums and 1KiB for the header", because it "was initially designed for keeping large amounts (like several thousands) of very big files"

MooseFS Master Server keeps all meta data in memory. Same issue as HDFS namenode.

Compared to Ceph

Ceph can be setup similar to SeaweedFS as a key->blob store. It is much more complicated, with the need to support layers on top of it. Here is a more detailed comparison

SeaweedFS has a centralized master group to look up free volumes, while Ceph uses hashing and metadata servers to locate its objects. Having a centralized master makes it easy to code and manage.

Ceph, like SeaweedFS, is based on the object store RADOS. Ceph is rather complicated with mixed reviews.

Ceph uses CRUSH hashing to automatically manage data placement, which is efficient to locate the data. But the data has to be placed according to the CRUSH algorithm. Any wrong configuration would cause data loss. Topology changes, such as adding new servers to increase capacity, will cause data migration with high IO cost to fit the CRUSH algorithm. SeaweedFS places data by assigning them to any writable volumes. If writes to one volume failed, just pick another volume to write. Adding more volumes is also as simple as it can be.

SeaweedFS is optimized for small files. Small files are stored as one continuous block of content, with at most 8 unused bytes between files. Small file access is O(1) disk read.

SeaweedFS Filer uses off-the-shelf stores, such as MySql, Postgres, Sqlite, Mongodb, Redis, Elastic Search, Cassandra, HBase, MemSql, TiDB, CockroachCB, Etcd, YDB, to manage file directories. These stores are proven, scalable, and easier to manage.

SeaweedFS comparable to Ceph advantage
Master MDS simpler
Volume OSD optimized for small files
Filer Ceph FS linearly scalable, Customizable, O(1) or O(logN)

Compared to MinIO, RustFS

Please note, as Apr 25, 2026 MinIO ceased development. It's strongly discouraged to use that unmaintained software with multiple security bugs. RustFS is a MinIO reimplementation in Rust, Apache 2.0 licensed and still developed, keeping MinIO's storage model down to a byte-compatible on-disk format. So the points below apply to both.

MinIO followed AWS S3 closely and was ideal for testing for S3 API. It had good UI, policies, versionings, etc. SeaweedFS is trying to catch up here.

The metadata are in simple files. Each file write incurs extra writes to the corresponding meta file, on every drive of the erasure set. Changing only tags or retention rewrites that meta file on all of them, so the write amplification does not shrink with object size.

There is no optimization for lots of small files. The files are simply stored as is to local disks. Plus the extra meta file and shards for erasure coding, it only amplifies the LOSF problem.

Multiple disk IO are needed to read one file. SeaweedFS has O(1) disk reads, even for erasure coded files.

Erasure coding is full-time. SeaweedFS uses replication on hot data for faster speed and optionally applies erasure coding on warm data.

No POSIX-like API support.

There are specific requirements on storage layout, which makes it hard to scale out and to maintain. An erasure set must be 2 to 16 drives and must divide the drive list symmetrically, and capacity grows or shrinks a whole pool at a time. In SeaweedFS, just start one volume server pointing to the master. That's all.

Back to TOC

Benchmark

Unscientific single-machine numbers from a MacBook with an SSD. weed benchmark, 1 million 1KB files, concurrency 16:

Requests per second p50 p99
Write 15,708 0.8 ms 2.6 ms
Random read 47,019 0.3 ms 0.7 ms

make benchmark runs warp mixed S3 traffic against a local weed server:

Mixed operations.
Operation: DELETE, 10%, Concurrency: 20, Ran 42s.
 * Throughput: 55.13 obj/s

Operation: GET, 45%, Concurrency: 20, Ran 42s.
 * Throughput: 2477.45 MiB/s, 247.75 obj/s

Operation: PUT, 15%, Concurrency: 20, Ran 42s.
 * Throughput: 825.85 MiB/s, 82.59 obj/s

Operation: STAT, 30%, Concurrency: 20, Ran 42s.
 * Throughput: 165.27 obj/s

Cluster Total: 3302.88 MiB/s, 550.51 obj/s over 43s.

Read throughput is bounded by the random read speed of the disks, and grows with every volume server added. More numbers, including multi-node, FUSE, and Hadoop, are in Benchmarks, S3 API Benchmark, FIO benchmark, and Independent Benchmarks.

Back to TOC

Enterprise

For enterprise users, please visit seaweedfs.com for the SeaweedFS Enterprise Edition, which has advanced features, including data recovery, self-healing storage, customizable erasure coding, EC vacuum and repair, etc.

Back to TOC

License

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

Back to TOC

Sponsors

Sponsor SeaweedFS via Patreon

SeaweedFS is an independent Apache-licensed open source project with its ongoing development made possible entirely thanks to the support of these awesome backers. If you'd like to grow SeaweedFS even stronger, please consider joining our sponsors on Patreon.

Your support will be really appreciated by me and other supporters!

Gold Sponsors

nodion piknik keepsec zyner

Back to TOC

Star History

Star History

Languages
Go 81.6%
Rust 11.2%
templ 2.4%
Java 1.5%
Python 0.9%
Other 2.1%