Hello there,
I am setting up a high volume media transcoding workflow using FFAStrans; and I could really use some guidance from the community. We handle a large amount of media files daily, and our current process is starting to show its limits in terms of efficiency and resource management.
We are running FFAStrans on a Windows Server 2019 system with 64 GB RAM and dual Intel Xeon processors.
Our workflow involves transcoding files from various formats into MP4 for web delivery and MXF for broadcast.
We are using network storage for file input/output, and we have a 1 Gbps network infrastructure.
What is the best approach to efficiently balance the workload across multiple CPU cores? Are there any specific configurations in FFAStrans I should be aware of for multi threading or parallel processing?
Occasionally; we run into transcoding errors due to corrupt source files or incorrect codecs. How do you typically manage job failures in FFAStrans?
Also, I have gone through this post; https://www.ffastrans.com/frm/forum/viewtopic.php?mlops=1276 which definitely helped me out a lot.
Are there recommended practices for optimizing workflow settings when handling high bitrate 4K content? I am particularly interested in reducing processing time while maintaining quality.
Thanks in advance for your help and assistance.
I Need Advice on Optimizing FFAStrans for High Volume Media Transcoding Workflow
-
- Posts: 2
- Joined: Mon Sep 09, 2024 7:32 am
Re: I Need Advice on Optimizing FFAStrans for High Volume Media Transcoding Workflow
Hi @roberrrttt1
your question title lets me think that next thing we read here is a link to some commercial system
It depends a lot on your workflow but generally you want to avoid filtering and avisynth (A/V Decoder), also you want to have 10Gbit Network. If there is no 10G infrastructure you could at least try to connect 2x1G and spread the workflow load onto the cables using different DNS names for your storages.
For optimizing single workflows you can send each one to us and we can check it for you.
Also, make sure h264 and h265 workflows do not have higher priority than normal.
your question title lets me think that next thing we read here is a link to some commercial system
It depends a lot on your workflow but generally you want to avoid filtering and avisynth (A/V Decoder), also you want to have 10Gbit Network. If there is no 10G infrastructure you could at least try to connect 2x1G and spread the workflow load onto the cables using different DNS names for your storages.
For optimizing single workflows you can send each one to us and we can check it for you.
Also, make sure h264 and h265 workflows do not have higher priority than normal.
Which processors exactly?roberrrttt1 wrote: ↑Wed Sep 18, 2024 9:34 am dual Intel Xeon processors.
What is the best approach to efficiently balance the workload across multiple CPU cores? Are there any specific configurations in FFAStrans I should be aware of for multi threading or parallel processing?
That is something unique for each and every user/workflow, there are so many ways to do handle errors... it is better that you imagine what is the best error handling scenario for you and then go and ask us how to come to that goalroberrrttt1 wrote: ↑Wed Sep 18, 2024 9:34 am Occasionally; we run into transcoding errors due to corrupt source files or incorrect codecs. How do you typically manage job failures in FFAStrans?
emcodem, wrapping since 2009 you got the rhyme?