#seaweedfs — Public Fediverse posts

So I have a #SeaweedFS what do I do with it

So I have a #SeaweedFS what do I do with it

Sorry for the storage outage few minutes ago. SeaweedFS seems to have another nasty bug: https://github.com/seaweedfs/seaweedfs/issues/10255

... which occurs from time to time 🙃

#seaweedfs #s3 #mastodon #metalheadclub

Sorry for the storage outage few minutes ago. SeaweedFS seems to have another nasty bug: https://github.com/seaweedfs/seaweedfs/issues/10255

... which occurs from time to time 🙃

#seaweedfs #s3 #mastodon #metalheadclub

Kasih kesempatan sekali lagi buat coba pakai erasure coding alias EC di warm storage.

Rusak ya rusak gpp, file-file ISO doang. Tinggal cari dan download lagi. Paling rugi waktu sama lonjakan CPU pas proses encode.

#SeaweedFS #ErasureCoding #ec

RE: https://mastodon.kodesumber.com/@seaweedfs_release_watcher/116792796158065636

All nodes updated to #SeaweedFS 4.35

4.35

What's Changed FUSE fix(command): preserve fuse option after writers by @7y-9 in #9972 test: add FUSE database load/durability/perf benchmark by @chrislusf in #9980 mount: tolerance-window write pattern detection for concurrent writeback by...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.35

#seaweedfs #s3 #objectstorage

RE: https://mastodon.kodesumber.com/@seaweedfs_release_watcher/116753228095040689

#seaweedfs cluster updated

4.34

What's Changed S3 s3: export per-bucket quota and read-only state metrics by @chrislusf in #9923 s3: validate copy source path segments by @chrislusf in #9929 [s3] validate indirect filer path inputs by @chrislusf in #9931 [s3]: preserve...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.34

#seaweedfs #s3 #objectstorage

@nopatience fortunately, my #mastodon doesn't have detected problems with my private #seaweedfs S3.

@nopatience fortunately, my #mastodon doesn't have detected problems with my private #seaweedfs S3.

RE: https://mastodon.kodesumber.com/@seaweedfs_release_watcher/116730547787348199

DONE, updated to v4.33

#seaweedfs

4.33

What's Changed S3 API fix(s3api): accept HTTP-date conditionals by @7y-9 in #9863 fix(s3api): drop ancestor directory markers from prefixed ListObjectVersions by @chrislusf in #9885 fix(s3api): clear stale object lock years by @7y-9 in...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.33

#seaweedfs #s3 #objectstorage

RE: https://mastodon.kodesumber.com/@seaweedfs_release_watcher/116713648967549972

Updated.

Tumben container enterprise-nya langsung release.

#seaweedfs

4.32

What's Changed Misc fix(http): check delete request errors before auth by @7y-9 in #9784 chore(weed/storage/needle): prune unused test functions by @alrs in #9812 [test] update docker image for s3test by @kmlebedev in #9811 security: hot-reload...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.32

#seaweedfs #s3 #objectstorage

Wow, that was quick! I reported a SeaweedFS bug before I went to bed.

Just got up from bed and a fix has been created! 😯 Now let's wait until a fixed version is released. I'll update metalhead.club asap.

https://github.com/seaweedfs/seaweedfs/issues/9695#event-25990705513

#seaweedfs #s3 #metalheadclub

Wow, that was quick! I reported a SeaweedFS bug before I went to bed.

Just got up from bed and a fix has been created! 😯 Now let's wait until a fixed version is released. I'll update metalhead.club asap.

https://github.com/seaweedfs/seaweedfs/issues/9695#event-25990705513

#seaweedfs #s3 #metalheadclub

Hm. Something broke our metalhead.club S3 server. Filed a bug report here: https://github.com/seaweedfs/seaweedfs/issues/9695

#seaweedfs #mastodon #s3 #metalheadclub

Hm. Something broke our metalhead.club S3 server. Filed a bug report here: https://github.com/seaweedfs/seaweedfs/issues/9695

#seaweedfs #mastodon #s3 #metalheadclub

@semurjengkol tanggung jg, bentar satu urusan lagi, tambah volume server buat #seaweedfs

@semurjengkol tanggung jg, bentar satu urusan lagi, tambah volume server buat #seaweedfs

Habe übrigens den SeaweedFS S3 Server für metalhead.club erfolgreich aktualisiert und dabei auch gleich auf "weed mini" umgestellt:

https://thomas-leister.de/mastodon-switch-minio-s3-seaweedfs/#update-vom-23-mai-2026

#s3 #seaweedfs #mastodon

Habe übrigens den SeaweedFS S3 Server für metalhead.club erfolgreich aktualisiert und dabei auch gleich auf "weed mini" umgestellt:

https://thomas-leister.de/mastodon-switch-minio-s3-seaweedfs/#update-vom-23-mai-2026

#s3 #seaweedfs #mastodon

@jwildeboer run #seaweedfs on #freebsd a couple of weeks … install and usage is fine

@jwildeboer run #seaweedfs on #freebsd a couple of weeks … install and usage is fine

SeaweedFS can be confusing sometimes, but Chris Lu, the project lead, seems to be quite helpful :)

#seaweedfs

SeaweedFS can be confusing sometimes, but Chris Lu, the project lead, seems to be quite helpful :)

#seaweedfs

Ich musste jetzt bei #SeaweedFS die Notbremse ziehen: die #Fnordon-Medien sind zu #Hetzner Object Storage umgezogen. Da habe ich weniger Kontrolle über die Daten, und es ist auch nur so halb selbst gehostet, aber plötzlich sind durch's Rüberkopieren aus 3 TB nur 600 GB geworden, und ich habe weniger KubePersistentVolumeFillingUp-Alarme.

#Mastoadmin

Ich musste jetzt bei #SeaweedFS die Notbremse ziehen: die #Fnordon-Medien sind zu #Hetzner Object Storage umgezogen. Da habe ich weniger Kontrolle über die Daten, und es ist auch nur so halb selbst gehostet, aber plötzlich sind durch's Rüberkopieren aus 3 TB nur 600 GB geworden, und ich habe weniger KubePersistentVolumeFillingUp-Alarme.

#Mastoadmin

Mit #SeaweedFS werde ich nicht glücklich, weil ich nicht verstehe, wie ich das Ding dazu bringe, den Platz von gelöschten Dateien wieder freizugeben.

Im Moment benötigen die ungefähr 450 GB Objekte von #Fnordon (größtenteils Cache der Medien anderer Server) insgesamt 2 TB Plattenplatz, Tendenz ständig steigend.

Irgendwas mache ich doch falsch.

#Mastoadmin

Mit #SeaweedFS werde ich nicht glücklich, weil ich nicht verstehe, wie ich das Ding dazu bringe, den Platz von gelöschten Dateien wieder freizugeben.

Im Moment benötigen die ungefähr 450 GB Objekte von #Fnordon (größtenteils Cache der Medien anderer Server) insgesamt 2 TB Plattenplatz, Tendenz ständig steigend.

Irgendwas mache ich doch falsch.

#Mastoadmin

What ended up working is using #Longhorn itself to backup my #Kubernetes volumes to my #S3 storage (#SeaweedFS), which works regardless of the volumes' access modes. This is fine, tho I still wish I could get #Velero working bcos in theory - it shd be better since it'd work with all storage rather than Longhorn alone... but yea honestly have no idea how to get it working with RWO volumes.

RE: https://sakurajima.social/notes/alt7i6brp4

What ended up working is using #Longhorn itself to backup my #Kubernetes volumes to my #S3 storage (#SeaweedFS), which works regardless of the volumes' access modes. This is fine, tho I still wish I could get #Velero working bcos in theory - it shd be better since it'd work with all storage rather than Longhorn alone... but yea honestly have no idea how to get it working with RWO volumes.

RE: https://sakurajima.social/notes/alt7i6brp4

@[email protected] i'd set up #seaweedfs as s3 storage, and tried to get velero working as means to backup my PVCs, but it ended up not working cos somehow, RWO #longhorn volumes of mine just fail to be backed up :/

@[email protected] i'd set up #seaweedfs as s3 storage, and tried to get velero working as means to backup my PVCs, but it ended up not working cos somehow, RWO #longhorn volumes of mine just fail to be backed up :/

I've set up #velero and #seaweedfs (as the #s3 storage) and yet backups of my pvcs are failing :( it seems that only the RWO vols are failing? but as far as im aware, velero/#kopia shdnt care/differentiate if it's RWO or not? any clue? underlying storage is #longhorn

#kubernetes #homelab

I've set up #velero and #seaweedfs (as the #s3 storage) and yet backups of my pvcs are failing :( it seems that only the RWO vols are failing? but as far as im aware, velero/#kopia shdnt care/differentiate if it's RWO or not? any clue? underlying storage is #longhorn

#kubernetes #homelab

Argh!

Debian Trixie maps /tmp directly to RAM—unlike previous versions <= Bookworm!

Since my Nginx cache is located in /tmp, my RAM filled up in no time after my Trixie upgrade. This, in turn, caused the OOM killer to target SeaweedFS, making the media storage inaccessible.

Sorry!

(But the problem has now been understood and fixed).

#metalheadclub #seaweedfs #mastodon #mastoadmin

Argh!

Debian Trixie maps /tmp directly to RAM—unlike previous versions <= Bookworm!

Since my Nginx cache is located in /tmp, my RAM filled up in no time after my Trixie upgrade. This, in turn, caused the OOM killer to target SeaweedFS, making the media storage inaccessible.

Sorry!

(But the problem has now been understood and fixed).

#metalheadclub #seaweedfs #mastodon #mastoadmin

#MinIO Is Done With #OpenSource, What Are Your Options?
MinIO is one of the most widely used #selfhosted #objectstorage solutions out there. It is #S3-compatible, lightweight, and runs as a single binary, integrating with pretty much everything in the #cloudnative stack.

Here are the three open source alternatives worth looking at. 👇
#SeaweedFS
#Garage
#RustFS

I have heard good things about Garage but haven't kicked the tires yet.
https://itsfoss.com/news/minio-moves-away-from-open-source/

#MinIO Is Done With #OpenSource, What Are Your Options?
MinIO is one of the most widely used #selfhosted #objectstorage solutions out there. It is #S3-compatible, lightweight, and runs as a single binary, integrating with pretty much everything in the #cloudnative stack.

Here are the three open source alternatives worth looking at. 👇
#SeaweedFS
#Garage
#RustFS

I have heard good things about Garage but haven't kicked the tires yet.
https://itsfoss.com/news/minio-moves-away-from-open-source/

От MinIO к SeaweedFS: опыт замены S3-хранилища

Есть такой опасный момент в инфраструктуре: когда все вроде бы работает, но трогать это лишний раз не хочется. Не потому что идеально. А потому что есть ощущение — если полезешь, станет хуже. В какой-то момент мы поймали себя на этом с MinIO.

https://habr.com/ru/companies/dbraincloud/articles/1028694/

#seaweedfs #minio #s3 #хранилище #кубернетес #kubernetes #контейнеризация #s3хранилище #инфраструктура #ceph

Achtung #Fnordon, jetzt rumpelt's gleich kurz.

Ich schalte den Object Storage von #MinIO auf #SeaweedFS um. Dadurch könnten vorübergehend Medien (Bilder/Videos) fehlen, die kommen aber alle wieder.

#Mastoadmin

Achtung #Fnordon, jetzt rumpelt's gleich kurz.

Ich schalte den Object Storage von #MinIO auf #SeaweedFS um. Dadurch könnten vorübergehend Medien (Bilder/Videos) fehlen, die kommen aber alle wieder.

#Mastoadmin

RE: https://mastodon.kodesumber.com/@seaweedfs_release_watcher/116436102190844395

Our containers have been updated and look stable.

#seaweedfs

"There is a new release of SeaweedFS. Happy updating your instance and don't forget to check the changelog."

4.21

What's Changed This is a packed release with many fixes and some new features! FUSE Mount now is 100% POSIX Compliant FUSE Mount added peer-to-peer file chunk sharing! For AI use case, added optimization for all-to-all parallel downloading from...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.21

#seaweedfs #s3 #objectstorage

"There is a new release of SeaweedFS. Happy updating your instance and don't forget to check the changelog."

4.2

What's Changed S3 API and IAM feat(iam): implement ListUserPolicies API action by @chrislusf in #8991 feat(iam): implement group inline policy actions by @chrislusf in #8992 fix(s3): don't count ErrNotFound as filer health failure in failover by...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.20

#seaweedfs #s3 #objectstorage

"There is a new release of SeaweedFS. Happy updating your instance and don't forget to check the changelog."

4.2

What's Changed S3 API and IAM feat(iam): implement ListUserPolicies API action by @chrislusf in #8991 feat(iam): implement group inline policy actions by @chrislusf in #8992 fix(s3): don't count ErrNotFound as filer health failure in failover by...

https://github.com/seaweedfs/seaweedfs/releases/tag/4.20

#seaweedfs #s3 #objectstorage

Expanding our #S3:

- #seaweedfs: 3 masters, 3 filers and 6 volume servers across 2 zones (datacenters)
- #garage: 2 different clusters, one of that consists of 6 nodes across 3 zones, the other is just 3 nodes in single zone.

But we still use S3 from cloud providers (mostly for archiving), because those are more flexible and cheaper on #capex.

Adding one more #seaweedfs volume server on secondary rack/datacenter.

I must watch does filer also write to other volume server on different rack/datacenter if I specify 001 replication setting. Gemini insists it wouldn't happen, but I have experienced it in the past.

Adding one more #seaweedfs volume server on secondary rack/datacenter.

I must watch does filer also write to other volume server on different rack/datacenter if I specify 001 replication setting. Gemini insists it wouldn't happen, but I have experienced it in the past.

Хаос в объектах: испытания Garage и SeaweedFS

С конца 2025 года известная многим, кто работает с object storage, система Minio начала издавать тревожные сигналы: разработчики полностью перестали принимать новые правки, прекратили обновлять Docker образ, убрали веб-интерфейс из опенсорсной версии, а с 13 февраля проект на GitHub полностью заморожен и стал архивом. Можно провести связь с покупкой Broadcom компании VMware, которой и принадлежит Minio, и последующим выжиманием денег из клиентов. Но мы не будем строить теории, есть вопрос интереснее: кто убережёт наши объекты, если не Minio? Давайте попробуем узнать на примере Garage и SeaweedFS. Мы будем бросать эти системы об стену, и смотреть, что получится на выходе. Так работает chaos testing. Наверное.

https://habr.com/ru/articles/1014226/

#seaweedfs #garage #objectstorage #s3 #minio

Как я искал замену MinIO S3 и написал свой S4 на Rust

В начале 2025 года компания MinIO в лице сооснователя Harshavardhana начала поэтапно сворачивать свою версию Community Edition. В феврале из open-source версии был вырезан веб-интерфейс администрирования - управление политиками, мониторинг, репликация, IAM - всё это переехало в коммерческий продукт AIStor с ценником от $96 000 в год. Пользователям оставили лишь базовый object browser и CLI-утилиту mc. В мае последовало удаление поддержки OIDC-аутентификации. В октябре MinIO прекратил публикацию Docker-образов и готовых бинарников - причём аккурат в момент раскрытия критической CVE-уязвимости. А в декабре 2025-го проект официально перешёл в режим maintenance mode: никаких новых фич, pull request'ы не принимаются, только точечные security-фиксы по усмотрению компании.

https://habr.com/ru/articles/1007060/

#minio #s3 #rust #seaweedfs #ceph #storage #rustfs #garage

@shlee @dansup
I am already a member of the #loops #discord ;)

When I think about it, it all boiled down to my lack of understanding how #s3 works.

Perhaps things are self-explaining when using #amazon s3 but I went with #minio at the first attempt and now with #seaweedfs (thx @thomas for the pointer due to his work on metalhead.club).

@shlee @dansup
I am already a member of the #loops #discord ;)

When I think about it, it all boiled down to my lack of understanding how #s3 works.

Perhaps things are self-explaining when using #amazon s3 but I went with #minio at the first attempt and now with #seaweedfs (thx @thomas for the pointer due to his work on metalhead.club).

So, I want to add an S3-compatible storage service local to my home lab (currently I'm using Backblaze). Basically I want it to run as a Docker service on my Synology NAS. It needs to run in the Synology because I want it to be external to my servers. I'm undecided between Garage or SeaweedFS.

Not MinIO. No, not Ceph. Not a VM. Not a Pi. Not in Kubernetes.

Anyone have/had that very particular setup and can share their experiences? Thanks in advance.

#HomeLab #Synology #Garage #SeaweedFS @homelab