Syncovery Support Forum

Hi,

I'm unable to find clear install instructions for a linux server. Ideally such instructions would include, for example, a wget command for obtaining syncovery and an authoritative list of dependancies.

For a server, download links seem inappropriate, and outdated references to dependancies leave me wondering...

Have I overlooked some "official" documentation.

Kind regards,

Steve

Hello,
here's an example for Debian systems:
There are no real dependencies. It uses libsqlite3 and libc, which all systems have anyway. OpenSSL or LibreSSL will be used for SSL connections, but if it's not there, Syncovery can still run and perform jobs that don't require SSL.

Thanks Tobias - that's perfect. I'll try an install in the morning.

The install on Ubuntu worked. Thanks.

For expedience I used the handy option to run the web GUI without SSL by adding ?nossl=1 to the URL. That got me to a small login screen. It asks for a username and password. I didn't see anywhere to set those up. My Ubuntu user login doesn't work.

How do I set up the username and password for the non-SSL login?

Hello,
the default login is
default
pass

It is mentioned many times on the download page.

When you're logged on, you can change it via the Program Settings button.

Thanks again for your help Tobias.

Tomorrow I'll see if I can set up a backup to S3.

As I mentioned in our email conversation I want to back up only files I've changed since installing the Ubuntu 20 image - i.e. I don't want to back up any of the original Ubuntu image files. So I'll need to figure out how to filter by date.

As far as directories to back up, I want to target / but exclude these directories: /dev, /run, /lost+found, /media, /misc, /mnt, /net, /proc, /sys, /var/run, /var/cache, /var/lock, /var/run,/ var/spool

Any tips appreciated.

Hello,
great. Setting up S3 should be pretty straightforward. Just click the Internet button for the right-hand side and change the protocol from FTP to S3.

Uploads to S3 lose the files's modification dates unless they are encoded into the filename. That's something to consider. The encoding option is on the Versioning tab sheet (Filename Encoding). So you need to decide if you can live with either the ugly encoded filenames on S3, or timestamps to be lost. I recommend using the filename encoding, because timestamps are important.

Of course all other metadata, such as permissions, are also lost, unless you choose to "compress each file individually" and choose permissions and extended attributes to be included in compressed files.

Finally, yes you can just specify / as the left-hand side base folder, and then put this into the Exclusion Masks field:
/dev;/run;/lost+found;/media;/misc;/mnt
/net;/proc;/sys;/var/run;/var/cache;/var/lock
/var/run;/var/spool

Newlines and semicolons are valid separators here.

Thanks again Tobias - really helpful, especially regarding preserving timestamps and metadata. Bravo.

I didn't see any mention of timestamp filtering (to avoid backing up the original Ubuntu image files). Is that because it's straightforward? In any case, I'll have a look...

Hello,
yes, the filters are under Masks & Filters -> File Age and Size.

Ok, found the filters - great. I've almost finished the setup. A couple things remain - you said:

metadata, such as permissions, are also lost, unless you choose to "compress each file individually" and choose permissions and extended attributes to be included in compressed files

1. I found the compress-each file option. I chose zip but note the sz option looks better. Are there any downsides (with S3 say)?

2. I'm unable to find with certainty the option to "choose permissions and extended attributes to be included in compressed files". Do you mean the last two options in the Permissions (see screenshot)? If so, can you please share your recommended checkbox ticks for a backup to S3. (I suppose the question is largely about the second checkbox in both sections.)