Hi @ all,
i have a problem with seahub after a restore. I’ve created a backup from my old seafile installation (seafile_server_5.1.3) and restored it to a new Ubuntu (16.04.1 LTS) installation. The seafile folder should be correct, because i’ve changed them with chmod to the new seafile user (and seafile group).
The seafile.sh scripts starts without problems and everything looks fine. If i try to start seahub.sh i get a error. Only if i start the seahub.sh as root everything seems fine. The logfiles are not very helpfull in this context. After this i’ve upgraded my seafile installation to seafile_server_6.0.3, but it appears the same error.
I do not use ngingx, apache, fastcgi or something like this and it does not help to switch the port.
Thanks and best regards
from the information given, we could only try to find out, what went wrong with your installation.
if you start seafile as the seafile user, what error message does the script produce?
What logfiles did you look into, and did they show?
try deleting /tmp/seahub_cache/* - this did it for me.
also check & see https://manual.seafile.com/faq/data-lost.html & https://manual.seafile.com/faq/
hope this helps.
your answer did it for me.
There was a problem with the access rights to seafile-directory/seafile-server-latest/runtime. They were set to root. The chown to my seafile user did it.
Thank you for your help.
Thanks for the hint on this. I’ve been fighting with having to start seahub as root for a few weeks without success. Tarsopilami and benschzilla forced me to take another look at file ownership in my seafile dirrectory, even though I was sure I had run chown on everything. There turned out to be serveral directories under seahub-data that were owned by other users. Once I changed those and removed /tmp/sehub_cache I was back in business, seahub now runs as the correct user.
Always run it with
-R to have it go through subdirs.