Flussonic

Flussonic V19 Resource Eating


#1

I’ve been using v18.06.1 for a long time since just before it HTTP was running fast (successive versions started to get worse). But when it came to replacing the SSD and reinstalling the OS, I decided to install the latest 19.07 and was amazed at how this version uses resources. Attached screenshots, with CPU and Scheduler utilization differences between versions. Resource growth is not with me alone, a colleague has also tested it on four servers and Scheduler utilization has doubled.

  1. http://www.ipix.lt/image/Vw9Cu
  2. http://www.ipix.lt/image/VwGDX

P.s. somehow only i noticed that WebOS did not want to accept AC3 tracks (does not even play channel) from Flussonic and they only corrected this in version 19.07 when I reported it but do not want to put AC3 fix in 18.06.1 package …


#2

Did you try this?

Here is our server:

  1. https://ibb.co/qYbjKWJ
  2. https://ibb.co/bXbv8Nf

We use about 261 streams (170 online) and about 400 users for now


#3

i have the same problem with flussonic update 18.02 was working fine
the new version have a lot of problems … resources eating , http receiving problems , transcoding ,dvr no sound … etc

evryday i hear something new

flussonic team better finding a solution than loosing customer …


#4

Yes , you are right about that. We have a big trouble a couple of weeks… All channels has interruptions, and its terrible. Here is some pictures. 746 clients (261 streams, online 136)
LOAD:
load
Traffic from server graphics with a lot of hight peacks
traffic

Real traffic from MAIN switch
realtraff

It``s pluged with optick on 10GB interface. Input and output in the same 10GB interface in different VLANS…

EDIT:

From today we split network traffic. Now our input is in different NIC 2x1g in bond with different vlans. Output is on 2x10G in bond, but we have exactly the same f…n problem! Anyone from Flussonic could you investigate and fix this isssue???

This is installed on ProLiant DL580 G7 with 4x(Execution Technology
8/8 cores; 16 threads ), and 32GB RAM. Is this not enought?


#5

hello,
i have same problem when i move from XC to flussonic
Screenshot_2019-10-02%20Flussonic%20Admin%20UI
i disable all and run only mpegts and same problem
the server is 20Gbps dedicated unmetered 24/7 + 8x1TB SSD + Dual Xeon E5-2650 + 256GB RAM
and run * streams: (586) * clients: 1409


#6

Is this with the new 19.10 V?


#7

yes V 19.9.1 i downgrade to 18.12 i think cpu better nowScreenshot_2019-10-03%20Flussonic%20Admin%20UI


#8

Single E5-2650 is not enough even for 100 channels. I have dual E5-2680 v3 and it can’t handle more than 300+ channels. Each channel needs about 2-2.5% CPU (E5-2680 v3). If CPU usage is more than 50% then flussonic can’t handle clients stable.


#9

Contact support or post here your root login/password


#10

It’s not just one user report this problem to check from server side. From couple of version till now you have a MAJOR problem with system resources… You need to figure out where is the problem, and what kind of task overload servers? For now in our servers and customers we DON`T have any interruptions, but the graph is terrible, and here is some LOG:
“Oct 11 09:29:16 transcoder flussonic[45719]: [os_mon] memory supervisor port (memsup): Erlang has closed
Oct 11 09:29:16 transcoder flussonic[45719]: [os_mon] cpu supervisor port (cpu_sup): Erlang has closed
Oct 11 09:29:18 transcoder flussonic[45719]: {“Kernel pid terminated”,application_controller,”{application_terminated,online,shutdown}"}
Oct 11 09:29:18 transcoder flussonic[45719]: Kernel pid terminated (application_controller) ({application_terminated,online,shutdown})
Oct 11 09:29:18 transcoder systemd[1]: flussonic.service: Main process exited, code=exited, status=1/FAILURE
Oct 11 09:29:19 transcoder control.erl[37544]: Flussonic is down
Oct 11 09:29:19 transcoder systemd[1]: flussonic.service: Control process exited, code=exited status=2
Oct 11 09:29:19 transcoder systemd[1]: flussonic.service: Unit entered failed state.
Oct 11 09:29:19 transcoder systemd[1]: flussonic.service: Failed with result ‘exit-code’.
Oct 11 09:29:19 transcoder systemd[1]: flussonic.service: Service hold-off time over, scheduling restart.
Oct 11 09:29:19 transcoder systemd[1]: Stopped Flussonic streaming server.
Oct 11 09:29:19 transcoder systemd[1]: Starting Flussonic streaming server…
Oct 11 09:29:32 transcoder flussonic[37853]: Preload took 9516 ms
Oct 11 09:29:38 transcoder systemd[1]: Started Flussonic streaming server."

Here is some graphics from today…

It’s not traffic problem and it`s not a server problem. Please try to fix because we plane to change flussonic with other alternative, because your price is not suitable with this problems