Frequent Repair Failed with HD Movies

[QuinRiva]

Version: 0.6.2
OS: Ubuntu 11.04 Server x64 (VM Guest on Server 2008 R2 Host)
Install-type: Linux Repository (run as daemon)
Firewall Software: None (that I'm aware of)
Are you using IPV6? I have not configure IPv6 for SABnzbd, but I do run a IPv6 DHCP/DNS servers.
Is the issue reproducible? yes 100%
USENET service: Newshosting.com

The majority of the movies that I have downloaded along with about 15% of the TV shows result in:

Code: Select all

Repair failed, not enough repair blocks (x blocks short)

where x is usually between 1 and 5.

As far as I can tell, nearly all files need to be repaired after downloading.

I also get this error occasionaly:

Code: Select all

When	Type	Warning
2011-05-28 06:25:04,550	ERROR:	[28/May/2011:06:25:04] Traceback (most recent call last):	 File "/usr/share/sabnzbdplus/cherrypy/_cpwsgi.py", line 79, in setapp	 s, h, b = self.get_response()	 File "/usr/share/sabnzbdplus/cherrypy/_cpwsgi.py", line 219, in get_response	 response = self.request.run(meth, path, qs, rproto, headers, rfile)	 File "/usr/share/sabnzbdplus/cherrypy/_cprequest.py", line 579, in run	 raise cherrypy.TimeoutError()	TimeoutError	
2011-05-28 06:25:04,511	ERROR:	[28/May/2011:06:25:04] Traceback (most recent call last):	 File "/usr/share/sabnzbdplus/cherrypy/_cpwsgi.py", line 79, in setapp	 s, h, b = self.get_response()	 File "/usr/share/sabnzbdplus/cherrypy/_cpwsgi.py", line 219, in get_response	 response = self.request.run(meth, path, qs, rproto, headers, rfile)	 File "/usr/share/sabnzbdplus/cherrypy/_cprequest.py", line 579, in run	 raise cherrypy.TimeoutError()	TimeoutError	

Re-downloading the file produces the same error.  How common is the error, and is there anything obvious that I'm doing wrong?
Also, how much RAM is recommend for SABnzbd

Further notes: the files were originally being downloaded to a windows share directory (privaleges were configured correctly as SABnzbd was being run through a Windows Domain registered user with full privileges for the shared directory).  This produced frequent CIFS error reporting unexepected SMB Signature.  I created a GPO disabling SMB signing on the the 2008 R2 Server and I changed the way I mounted the shared directories and the CIFS problem seems to be resolved. Shares are now mounted using VMWare's shared directories (hgfs).

[shypike]

The timeout error is unrelated, it means the system is so busy that it
cannot handle UI requests.

After a download has completed (and failed)
are all expected files, including par2 files in?
Did you try to check with a program like QuickCheck (in Windows)?

[QuinRiva]

I'm not sure what QuickCheck is but I ran QuickPar (in Windows) and got the following result:




There just seems like I'm getting loads of missing blocks, surely that can't be normal?

[pobox]

Newshosting completion is spotty 500-700 days old.  I suggest adding backup servers paid for by block.

[shypike]

The more well-known Usenet providers are frequently hit by DMCA notices.
Some remove everything, some just enough so that you won't be able to complete.
Often sites like binsearch.info show a missing percentage.
However, once the servers have been indexed a site like binsearch no longer
keeps track of what happens on specific servers.

I'm not saying that is the issue here, but it's a factor to keep in mind.

SABnzbd has its issues too, but the basic downloading is solid.

[QuinRiva]

Newshosting completion is spotty 500-700 days old.  I suggest adding backup servers paid for by block.

What do you mean by adding back up servers?  Can you suggest a better service than Newshosting?