Flix Cloud is experiencing slow job processing & response times. We’re looking into it & will post reports as we learn more. Sorry for the inconvenience.

[UPDATE 11:52AM EST] We’re back up but slow. Accepting job requests, but with a limited number of workers. Hopefully be able to ramp up workers through the afternoon.

[UPDATE 11:43AM EST] We found multiple instanced of our queue scripts running on the DB server. These may have contributed to high CPU load.

[UPDATE 11:27AM EST] Our suspicion is that there is an outer_join query somewhere in the system keeps throttling the database. We had to block incoming jobs to try & locate the problem query.

A number of customers have opened support cases regarding sftp file transfer errors and failures.

Most of the time the failure is due to file paths. When using sftp for input & output file locations in your job requests, the file path you specify must be the full, absolute path to the file, not a location relative to the home directory of the sftp user you connected as.

For example, if you connect to your server as user sftpuser, instead of a shorthand location like sftp://f1.vidfiles.com/output.flv, you must specify sftp://f1.vidfiles.com/home/sftpuser/outfile.flv

You do not need to do this for regular FTP. You case use relative paths with regular FTP.

This week we switched from ffmpeg to mediainfo for incoming video inspection. Mediainfo is a very nice tool, and failed inspections have all but disappeared since we deployed it in Flix Cloud.