Architecture for Predicting Live Video Transcoding Performance on Docker Containers

Pekka Pääkkönen; Antti Heikkinen; Tommi Aihkisalo

doi:10.1109/SCC.2018.00016

Architecture for Predicting Live Video Transcoding Performance on Docker Containers

Pekka Pääkkönen, Antti Heikkinen, Tommi Aihkisalo

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

3 Citations (Scopus)

Abstract

Video can be streamed live with differentapplications (e.g. YouTube Live, Periscope). Typically, thevideo content is adapted for end users based on receivingclient’s capabilities, and network bandwidth. The adaptation isrealized with different video representations, which are createdby transcoding the original video content. When video isstreamed live, transcoding has to be completed within real timeconstraints, which is a computationally demanding process.Particularly, live transcoding should be enabled efficiently by acontent distributor to minimize resource provisioning costs.The contribution of this paper is an architecture for predictinglive video transcoding performance on a Docker-basedplatform. Particularly, cloud resource management for livevideo transcoding has been focused on. A model was trainedbased on measurements in different transcodingconfigurations. Offline evaluation results indicate that livetranscoding speed or CPU usage can be predicted with 3-8 %accuracy. When video is transcoded on virtual machines basedon predictions in a prototype system (live), live transcodingspeed prediction accuracy is within a similar range as theoffline performance, but worse for CPU usage prediction (5-15%). In most cases the specified range for transcoding speedand CPU usage can be achieved at least with a precision of 76%.

Original language	English
Title of host publication	2018 IEEE International Conference on Services Computing, SCC 2018
Subtitle of host publication	Part of the 2018 IEEE World Congress on Services
Publisher	IEEE Institute of Electrical and Electronic Engineers
Pages	65-72
Number of pages	8
ISBN (Electronic)	978-1-5386-7250-1
ISBN (Print)	978-1-5386-7251-8
DOIs	https://doi.org/10.1109/SCC.2018.00016
Publication status	Published - 1 Jul 2018
MoE publication type	Not Eligible
Event	IEEE International Conference on Services Computing, SCC 2018 - San Francisco, United States Duration: 2 Jul 2018 → 7 Jul 2018

Conference

Conference	IEEE International Conference on Services Computing, SCC 2018
Abbreviated title	SCC 2018
Country/Territory	United States
City	San Francisco
Period	2/07/18 → 7/07/18

Keywords

video transcoding
FFmpeg
Rancher
Cassandra
Docker
Random forest
Prometheus

Access to Document

10.1109/SCC.2018.00016

Cite this

Pääkkönen, P., Heikkinen, A., & Aihkisalo, T. (2018). Architecture for Predicting Live Video Transcoding Performance on Docker Containers. In 2018 IEEE International Conference on Services Computing, SCC 2018: Part of the 2018 IEEE World Congress on Services (pp. 65-72). Article 8456402 IEEE Institute of Electrical and Electronic Engineers. https://doi.org/10.1109/SCC.2018.00016

@inproceedings{ebac1b68b55f47588e011a6a241d09ce,

title = "Architecture for Predicting Live Video Transcoding Performance on Docker Containers",

abstract = "Video can be streamed live with differentapplications (e.g. YouTube Live, Periscope). Typically, thevideo content is adapted for end users based on receivingclient{\textquoteright}s capabilities, and network bandwidth. The adaptation isrealized with different video representations, which are createdby transcoding the original video content. When video isstreamed live, transcoding has to be completed within real timeconstraints, which is a computationally demanding process.Particularly, live transcoding should be enabled efficiently by acontent distributor to minimize resource provisioning costs.The contribution of this paper is an architecture for predictinglive video transcoding performance on a Docker-basedplatform. Particularly, cloud resource management for livevideo transcoding has been focused on. A model was trainedbased on measurements in different transcodingconfigurations. Offline evaluation results indicate that livetranscoding speed or CPU usage can be predicted with 3-8 %accuracy. When video is transcoded on virtual machines basedon predictions in a prototype system (live), live transcodingspeed prediction accuracy is within a similar range as theoffline performance, but worse for CPU usage prediction (5-15%). In most cases the specified range for transcoding speedand CPU usage can be achieved at least with a precision of 76%.",

keywords = "video transcoding, FFmpeg, Rancher, Cassandra, Docker, Random forest, Prometheus",

author = "Pekka P{\"a}{\"a}kk{\"o}nen and Antti Heikkinen and Tommi Aihkisalo",

year = "2018",

month = jul,

day = "1",

doi = "10.1109/SCC.2018.00016",

language = "English",

isbn = "978-1-5386-7251-8 ",

pages = "65--72",

booktitle = "2018 IEEE International Conference on Services Computing, SCC 2018",

publisher = "IEEE Institute of Electrical and Electronic Engineers",

address = "United States",

note = "IEEE International Conference on Services Computing, SCC 2018, SCC 2018 ; Conference date: 02-07-2018 Through 07-07-2018",

}

Pääkkönen, P , Heikkinen, A & Aihkisalo, T 2018, Architecture for Predicting Live Video Transcoding Performance on Docker Containers. in 2018 IEEE International Conference on Services Computing, SCC 2018: Part of the 2018 IEEE World Congress on Services., 8456402, IEEE Institute of Electrical and Electronic Engineers, pp. 65-72, IEEE International Conference on Services Computing, SCC 2018, San Francisco, California, United States, 2/07/18. https://doi.org/10.1109/SCC.2018.00016

Architecture for Predicting Live Video Transcoding Performance on Docker Containers. / Pääkkönen, Pekka ; Heikkinen, Antti ; Aihkisalo, Tommi.
2018 IEEE International Conference on Services Computing, SCC 2018: Part of the 2018 IEEE World Congress on Services. IEEE Institute of Electrical and Electronic Engineers, 2018. p. 65-72 8456402.

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Architecture for Predicting Live Video Transcoding Performance on Docker Containers

AU - Pääkkönen, Pekka

AU - Heikkinen, Antti

AU - Aihkisalo, Tommi

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Video can be streamed live with differentapplications (e.g. YouTube Live, Periscope). Typically, thevideo content is adapted for end users based on receivingclient’s capabilities, and network bandwidth. The adaptation isrealized with different video representations, which are createdby transcoding the original video content. When video isstreamed live, transcoding has to be completed within real timeconstraints, which is a computationally demanding process.Particularly, live transcoding should be enabled efficiently by acontent distributor to minimize resource provisioning costs.The contribution of this paper is an architecture for predictinglive video transcoding performance on a Docker-basedplatform. Particularly, cloud resource management for livevideo transcoding has been focused on. A model was trainedbased on measurements in different transcodingconfigurations. Offline evaluation results indicate that livetranscoding speed or CPU usage can be predicted with 3-8 %accuracy. When video is transcoded on virtual machines basedon predictions in a prototype system (live), live transcodingspeed prediction accuracy is within a similar range as theoffline performance, but worse for CPU usage prediction (5-15%). In most cases the specified range for transcoding speedand CPU usage can be achieved at least with a precision of 76%.

AB - Video can be streamed live with differentapplications (e.g. YouTube Live, Periscope). Typically, thevideo content is adapted for end users based on receivingclient’s capabilities, and network bandwidth. The adaptation isrealized with different video representations, which are createdby transcoding the original video content. When video isstreamed live, transcoding has to be completed within real timeconstraints, which is a computationally demanding process.Particularly, live transcoding should be enabled efficiently by acontent distributor to minimize resource provisioning costs.The contribution of this paper is an architecture for predictinglive video transcoding performance on a Docker-basedplatform. Particularly, cloud resource management for livevideo transcoding has been focused on. A model was trainedbased on measurements in different transcodingconfigurations. Offline evaluation results indicate that livetranscoding speed or CPU usage can be predicted with 3-8 %accuracy. When video is transcoded on virtual machines basedon predictions in a prototype system (live), live transcodingspeed prediction accuracy is within a similar range as theoffline performance, but worse for CPU usage prediction (5-15%). In most cases the specified range for transcoding speedand CPU usage can be achieved at least with a precision of 76%.

KW - video transcoding

KW - FFmpeg

KW - Rancher

KW - Cassandra

KW - Docker

KW - Random forest

KW - Prometheus

UR - http://www.scopus.com/inward/record.url?scp=85054004346&partnerID=8YFLogxK

U2 - 10.1109/SCC.2018.00016

DO - 10.1109/SCC.2018.00016

M3 - Conference article in proceedings

SN - 978-1-5386-7251-8

SP - 65

EP - 72

BT - 2018 IEEE International Conference on Services Computing, SCC 2018

PB - IEEE Institute of Electrical and Electronic Engineers

T2 - IEEE International Conference on Services Computing, SCC 2018

Y2 - 2 July 2018 through 7 July 2018

ER -

Architecture for Predicting Live Video Transcoding Performance on Docker Containers

Abstract

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this