Failed to start seafile server

yep, it sounds should be like that. But there is no bin folder in the seafile folder.

It should be located in /opt/share/seafile/seafile-server/seafile

will check it later… on My way to the office

No problem.

Nope, there’s nothing in that folder…

Here’s the information:

[K3 /tmp/share/sda1/opt/share/seafile/seafile-server/seafile]$ ls -a
.   ..
[K3 /tmp/share/sda1/opt/share/seafile/seafile-server/seafile]$

The seafile-controller is really located in /opt/bin (the path was generated during installation):

[K3 /tmp/share/sda1/opt/share/seafile/seafile-server/seafile]$ ls /opt/bin
bash                    locale.new              rbash
ccnet-init              localedef.new           seaf-fsck
ccnet-server            logger                  seaf-migrate
chardetect              opkg                    seaf-server
createfontdatachunk.py  painter.py              seaf-server-init
django-admin.py         pilconvert.py           seafile-admin
easy_install            pildriver.py            **seafile-controller**(I added stars here...)
easy_install-2.7        pilfile.py              seafserv-gc
enhancer.py             pilfont.py              sqlite3
explode.py              pilprint.py             sudo
find                    pkill                   thresholder.py
gifmaker.py             player.py               tificc
gunicorn                psicc                   transicc
gunicorn_paster         python                  viewer.py
jpgicc                  python2                 vim
linkicc                 python2.7               xargs

Anyway, the seafile.sh could start seafile-controller correctly, so I don’t think the path has anything to do with the failure.

And I found the port forwarding is not working at all… The K3 router doesn’t allow SSH connection from the WAN port.

Yes, you could theoretically start the server from anywhere. The issue becomes the version at that point. The update scripts may not update that one in the future, and since your seafile.sh script you are using is hard coding some of it, the paths may be getting screwed up somewhere since there are dozens of scripts that Seafile runs, and that path could be passed to other scripts. The manual states that the seafile.sh scripts should be run from the Seafile folder.

What I’m wanting to do is compare the scripts in the Seafile folder with the scripts in /opt/bin and also taking a look at permissions as well as versions. I also want to manually run pgrep and see what it returns since that’s what appears to be throwing the ā€œseafile failed to startā€ error. Yes, it’s in the running processes, but why is it throwing that error? For some reason, the code is reporting that seafile-controller is not running even though it is. So, there are several things we need to look at to narrow this down.

I just noticed in your controller.log that the seafile pid is being executed from /var/opt/run, which is another folder outside the seafile folder from which something is executed. I would like to also take a look at the script you used to install seafile and compare it to the script from the CE edition obtained from the seafile web site.

i just installed seafile using ā€˜opkg install seafile-server’, then everything goes on automatically…like apt-get install on Ubuntu.

I’ve put the teamviewer online, hoping this could solve the NAT thing… I sent the ID and password in PM.

total Junk. why not just put the error in the logs and call it a day. would save everybody months of browsing forums