I ran a garbage collection and it removed many unreachable blocks (over 100GB). The library size within the admin web interface is shown as around 370GB whereas the actual size of the library on the file system (in seafile-data/storage/blocks/<id> using du -sch * is around 310GB after the garbage collection.
This results in several questions:
- Am I looking at the right place for the library size on the file system in
seafile-data/storage/blocks?
- Could it be that the garbage collection removed too many blocks?
- Is it save to start the client for synchronizing?
- In case something is wrong here, where can I look for the problem?
1 Like
Remember that seafile devides data in little chunks under the hood and that chunks with the same information are only saved once (deduplication) maybe the Webinterface Displays the data without dedup .
E.g. you have multiple windows isos on the disk
Certain data will only be effectively written to disk, the duplicates will only get a link to the original chunk location.
Just my assumption to your problem.
1 Like
@mulmer This is what i would think.
I already had libraries with 1 TB but they just used around 500 GB on storage.
So nothing wrong here.
Thanks for your feedback.
The deduplication might be the reason, although I am still a bit skeptic. There are mainly photos in that library where I would think it’s quite unlikely to have identical chunks.
I will check whether there are duplicate photos or other files in that library which might be the reason.
1 Like