Question about what's happening

Seafile Server
1

I have been using Nextcloud for about a year, and was recently advised that my needs would be better server by Seafile, so I’m trying it out.

I use my Linux server to synchronise music (mostly flac files), ebooks (mostly epubs), and miscellaneous other things (documents, photos, etc) between 3 laptops, and often access them with an android phone and an android pad/tablet. So, if, for example, I get a new ebook, I only need to copy it to one laptop and it synchronises automatically to the server, and then to the other laptops. It then can be access from the server by the android devices.

So, I installed Seafile on an Ubuntu 18.04 server by running the (rather excellent) installation script. All good.

I then loaded the data onto the server by using the Linux Seafile client app’s synchronise function on one of my laptops (running Debian 10). This took a few hours to complete. I then connected a 2nd laptop (also Linux - I don’t do Windows), which already had the data on it (earlier synchronised using Nextcloud, which I had disabled for this exercise). I was expecting the second laptop to run quite quickly as the data was the same as that on the first laptop. But, an hour later, it was still “indexing”. I noticed that the date/timestamps on the second laptop had changed, so it seems to be over-writing the laptop’s existing data with that on the server. But… I don’t see much network I/O, so it can’t be physically downloading the files.

What is going on? The second laptop looked like it would also take a few hours to complete (I aborted the process after about an hour). Why would it need to replace data that’s already there? Or… is it replacing it? During the initial server load, I monitored the network traffic and it was going at about 50Mb/s. The 2nd laptop - which already had the data - was running at a few Kb/s.

I’m rather confused :slight_smile:

Thanks so much!

2

I’d advise you read https://seafile.gitbook.io/seafile-server-manual/developing/sync-algorithm to get an idea of how seafile works, most likely the second laptop was checking if the hashes for the files on the directory matched the ones on the server. You did not specify if the disk IO stats you mentioned are from read, writes or the sum of them.

As for the modified date changes, I think seafile also syncs those with the server given that you had already uploaded those files, but I’m not sure.

1 Like
3

Thanks so much for you’re reply. I’ve got it up and running, and I’m quite impressed. I’m thinking of upgrading to the Pro edition as I only have 2 users: admin and guest. I’m just not sure that I need it.