How-to to setup Grafana dashboards to monitor Jitsi, my comprehensive tutorial for the beginner

Thanks, anyway @Woodworker_Life was enough clear to express why it was removed :slight_smile:

as @Woodworker_Life already stated, the parameter [[inputs.cpu]] is independent from jitsi - so simply use this.

Yes mee too everything work fine except CPU and Memory.
Any help with this two points?.

Thanks

Hi @mmgnservices,
No offense, but please read carefully. The solution is already twice in this thread.
Add a file /etc/telegraf/telegraf.d/system.conf

###############################################################################
#                                  INPUTS                                     #
###############################################################################

[[inputs.cpu]]
## Whether to report per-cpu stats or not
    ## percpu = true
    ## Whether to report total system cpu stats or not
    totalcpu = true

[[inputs.mem]]

###############################################################################
#                                  OUTPUTS                                    #
###############################################################################

[[outputs.influxdb]]
    urls = ["http://localhost:8086"]
    database = "jitsi2"
    timeout = "0s"
    retention_policy = ""

Hey there,
thanks for this guide. Helped me a lot.
My Grafana shows some fancy numbers but I’m missing some graphs.
All “commom” values are shown as no. of channels, no. of participants and so on. But none of the graphs is showing any lines. They’re only hovered by the message “data outside time range”. I tired to find any solution but it seems some sort of format problem. I made a short dump on my console using curl -v … and see that the timestamp is about 2 hrs behind local timezone. Any ideas is this might be the source of this issue?
Data for the graphs is provided. I can see some correct values below the graphs in the table.
I haven’t found any similar problem on the web, yet. Or I might have overseen it.
Any suggestions? Any solutions?
Thanks for your support,
Dirk

hi, did you get it running? i dont get how to do it, i have no clue where to configure these things… can you help me? just to be sure, you have 2 seperated servers and each server should be monitored ?

Hi all great thread
Just a thought, with influxDB ver 2, the dashboard within InfluxDB might also serve the purpose without the need of Grafana. I actually have skipped InfluxDB altogether though. My setup is Telegraf–Prometheus–Grafana. Also its just monitoring a single JVB
Vivek

How we can define two jvb1 and jvb 2 and monitoring them individually in Grafana.

vim /etc/telegraf/telegraf.conf
[global_tags]

###############################################################################

GLOBAL

###############################################################################

[agent]
interval = “10s”
debug = false
hostname = “jitsi_host”
round_interval = true
flush_interval = “10s”
flush_jitter = “0s”
collection_jitter = “0s”
metric_batch_size = 1000
metric_buffer_limit = 10000
quiet = false
logfile = “”
omit_hostname = false
#cpu_usage=“0”
#[[inputs.cpu]]
#percpu = true
#totalcpu = true
#collect_cpu_time = true
#report_active = true
#cpu_usage = “0s”

sudo nano /etc/telegraf/telegraf.d/jitsi.conf

###############################################################################

INPUTS

###############################################################################

[[inputs.http]]
name_override = “jitsi_stats”
urls = [
“bridge1.xx:8080/metrics”,
“bridge2.xx:8080/metrics”
]

   data_format = "json"

###############################################################################

OUTPUTS

###############################################################################

[[outputs.influxdb]]
urls = [“localhost:8086”]
database = “telegraf2020”
timeout = “0s”
retention_policy = “”

3 Likes

Hello!

I followed the tutorial but I’m in step when I have to login to grafana with admin user, but I don’t know the information to login and the tutorial only tells this:

Open Grafana in browser and prepare your admin account

My Jitsi and Grafana server are in the same server, so, the address is localhost:3000.

Any ideas?

Thank you, now I access, no problem!

So, when I create the database, when I do a “Save & Test”, give me this error => Network Error: Bad Gateway (502).

What happened?

Hi, did you solve this problem?

Previously my installation work fine, and now i want to move the grafana server to another machine and i’m also getting this error when creating database.

I can manually get the output of: curl -v http://jitsi_server:8080/colibri/stats on the new grafana machine, but still get the 502 bad gateway error.

Hello @Access_Lab

No, I didn’t solve my problem and nobody answer here for try to solve this :frowning:

Hi @Ferran_Munoz,

Try this, it solves my problem, it seems that I missed this step when adding an influxDB datasource:

URL: http://localhost:8086 --> you have to type instead of use the default settings (I’ve missed this step previously and it cause 502 bad gateway error)

Good luck with your settings

1 Like

Thanks a lot for the tutorial ! It helped a lot.

I have a question. Did anybody find a good way to visualize the information in
“conference_sizes”, “conferences_by_video_senders” and “conferences_by_audio_senders” ?

It seems that is really a wealth of information in there, it would be very valuable for us to know the details of the conference sizes, and also how many people turn off their webcam or mike, but I really struggle to show it in grafana. Any ideas ?

Thanks @Access_Lab, Your solution worked for me :relaxed:

I had this problem and I have fixed it. I have configured this with kubernetes? How about your environment?

Hello @Woodworker_Life and all!

By any chance does anyone knows if the author of this dashboard hangs around Jitsi Discourse?

Thanks in advance

@Woodworker_Life Very interesting and useful contribution. Thanks a lot.
Best wishes for your work and life.

Sudhir Gandotra
Best Book Buddies

Hi there,

we have successfully managed to monitor our 3 videobridges by adding the stats of them in separate influx databases via telegraf.
Now what we would really like to see is some panel with aggregated data from all videobridges.
For example we would love to have a panel with the realtme amount of participants or confereces running.
Anyone has any idea or allready implemented this somehow?

Thanks
Sascha

Hi, did you manage to monitor multi JVBs with grafana? i found that there number of conf is just showing last record coming from any of JVB that is not correct. appreciated if you share your experience.

1 Like