Hi Seafile fans.
I’m using SeaDrive on Linux over terminal.
Does anyone have an idea how could it be that some files are visible in the SeaDrive folder using terminal, while they are not visible online in the seafile?
Is there any way to force the sync between the both?
I suspect they haven’t been uploaded, yet.
They are uploaded. I rolled back one older zfs snapshot and it started to fetch some files. But some files didn’t fetch. So there is a difference. Is there any file that stores information about already uploaded files, which, if deleted, would initiate the process of comparison between already uploaded and not uploaded files on the web?
I think there is an option to resync a library.
I don’t see anything in the manual. Do you have an idea how?
SeaDrive does not have a resync function. The Seafile Sync Client has.
If you don’t see all files, the most likely reason is that your library is too large. How many files do you have in the library? (You can check that via System Admin in the web interface.)
This library to large thing is the worst change ever.
I am not even talking about the max_file_count parameter introduced in Seafile Server 7.1.16 and 8.0.3. This can be configured.
My personal experience is that SeaDrive sometimes causes trouble with libraries holding many many thousand files. I am explicitly vague regarding the threshold because I have seen SeaDrive work just fine with libraries with in excess of 500.000 files. On the other hand, I have seen SeaDrive not syncing libraries correctly with 250.000 files.
So, should I use Seafile Client on my server instead to upload lots of data to Seafile? I have more than 1 million files.
It is recommended not to store more than 100.000 files per library. This is the official recommendation and it is independent of the client.
As I said above: Seafile does not know a hard limit. I’ve seen libraries with close to a million files which didn’t cause any trouble. But there is no guarantee that it stays this way and the moment you experience problems, I’d tell you: Tough luck.
My recommendation: Do not even try to store 1M files in a library. Think about ways how to break up your data in a smart way. Based on experience, this is usually not so hard. You’ll be rewarded with a Seafile Server that is very stable and reliable.