[seafile-pro-server-6.0.13] PDF --> Internal error


I looked through the posts, but unfortunately I did not find the answer to my problem: I can’t preview the pdf file in Seafile.
I configured this functionality according to the Seafile guide.
When I try to open pdf file, I receives “Internal error”.

The only log that appears in system is seahub.log:
2017-05-22 09:24:16,127 [ERROR] root:1087 prepare_converted_html failed to add_office_convert_task:
Traceback (most recent call last):
File “/data/haiwen/seafile-pro-server-6.0.13/seahub/seahub/utils/init.py”, line 1085, in prepare_converted_html
add_office_convert_task(obj_id, doctype, raw_path)
File “/data/haiwen/seafile-pro-server-6.0.13/seahub/seahub/utils/init.py”, line 997, in real_func
return delegate_func(*args)
File “/data/haiwen/seafile-pro-server-6.0.13/seahub/seahub/utils/init.py”, line 1013, in delegate_add_office_convert_task
ret = do_urlopen(url, data=data, headers=headers).read()
File “/data/haiwen/seafile-pro-server-6.0.13/seahub/seahub/utils/init.py”, line 1214, in do_urlopen
ret = urllib2.urlopen(req)
File “/usr/lib/python2.7/urllib2.py”, line 154, in urlopen
return opener.open(url, data, timeout)
File “/usr/lib/python2.7/urllib2.py”, line 421, in open
protocol = req.get_type()
File “/usr/lib/python2.7/urllib2.py”, line 283, in get_type
raise ValueError, “unknown url type: %s” % self.__original
ValueError: unknown url type: /office-convert/internal/add-task/

System info:

  • Ubuntu 16.04.2 LTS
  • Seafile-pro-server-6.0.13
  • two node cluster + haproxy + keepalived
  • libreoffice
  • poppler-utils 0.41

Best regards,
Krzysztof Wadowka.

I believe this is a configuration issue. Since you have only two nodes, your configuration may be somehow different from what’s documented in https://manual.seafile.com/deploy_pro/enable_search_and_background_tasks_in_a_cluster.html . We usually recommend to have a separate node as background node. But I think with some care, you can also set it up with two nodes.

Can you check this documentation again and find any problems in your configuration?

Hi Jonathan,

thank you for quick reply.

I would like to avoid configuration background node. So far I configured the service on each node as in the following link:

I want to get a fully universal configuration of Seafilepro instance (for the purpose of adding a new nodes to cluster in future).

Best regards,


We do recommend you to follow the documentation structure. We don’t assume a node to be both a front end and background node in our cluster design. So there could be hidden problems. Furthermore, you should not use the same configuration to start front end and background node. So my suggestion is to use a separate node for background tasks and make a configuration for front-end nodes. You can use that configuration to start more front-end nodes in the future.