@steschuser , your comment on using glances with over 20 machines, has spurred my interest. Could you please enlighten me on the best setup. I want to monitor many virtual machines - I intend to use the results to trigger some algorithms implementation. so am I suppose to run
glance -w on all these machines? Then to get the result, can I just call API end points to get them? Also, how can I ensure glances are running say like 24/7 on the machines. If you use influx, does it overwrite results or it just stores them at the different intervals?
@aadewojo
you can start glances via systemd
you can start glances via systemd
[Unit]
Description=Glances
After=network.target influxd.service
[Service]
ExecStart=/usr/bin/glances --quiet --export-influxdb -t {{influxdb_intervall | default("30")}}
Restart=on-failure
RestartSec=30s
TimeoutSec=30s
[Install]
WantedBy=multi-user.target
influx is a time-based database, so nothing gets overwritten.
and each machine has its own prefix
and you can easily monitor systemd and have it restart on failure
And the corresponding glances configuration is like this
[influxdb]
# Configuration for the --export-influxdb option
# https://influxdb.com/
host={{influxdb_server | default("localhost")}}
port=8086
user={{influxdb_username | default("root")}}
password={{influxdb_password | default("root")}}
db=glances
prefix={{ansible_hostname}}
tags=system:{{ansible_hostname}}
@aadewojo
hello again,
I have no experience with influx cloud
I installed infludb and grafana on a single host influxdb_server
glances will transport the data there
https://www.tecmint.com/install-glances-influxdb-grafana-to-monitor-centos-7/
I used this grafana dashboard to display the grafana data https://grafana.com/grafana/dashboards/2387
hello again,
I have no experience with influx cloud
I installed infludb and grafana on a single host influxdb_server
glances will transport the data there
https://www.tecmint.com/install-glances-influxdb-grafana-to-monitor-centos-7/
I used this grafana dashboard to display the grafana data https://grafana.com/grafana/dashboards/2387
However, I have got a question please. I can't see any metrics on grafana dashboard. I tried to investigate and found a query used to inspect "CPU Core" as
{
"request": {
"method": "GET",
"url": "api/datasources/proxy/2/query",
"params": {
"db": "glances",
"q": "SELECT mean(\"().cpu_percent\") FROM \"localhost.docker\" WHERE time >= now() - 15m GROUP BY time(500ms) fill(none);SELECT mean(\"().memory_usage\") FROM \"localhost.docker\" WHERE time >= now() - 15m GROUP BY time(500ms) fill(none)",
"epoch": "ms"
},
"data": null,
"precision": "ms"
},
"response": {
"results": [
{
"statement_id": 0
},
{
"statement_id": 1
}
]
}
}
I feel it shouldn't be localhost.docker
Could you @steschuser or anyone shed some light on this please?
Hi, I'm running glances with export influxdb and it seems to be working (I can see the database and tables). Now in grafana I added the dashboard but all graphics are with "no data points". There is no hosts in host list but I know that grafana can read the data because the interface and disk list have the names if I manually set the host.
I setup the log and the only problem are related to docker and gpu.
2020-06-29 16:50:21,113 -- ERROR -- docker plugin - Cannot get Docker version (('Connection aborted.', FileNotFoundError(2, 'No such file or directory')))
2020-06-29 16:50:21,436 -- ERROR -- Cannot export hank.docker stats to InfluxDB (400: {"error":"unable to parse 'hank.docker ': invalid field format"}
)
2020-06-29 16:50:21,445 -- ERROR -- Cannot export hank.gpu stats to InfluxDB (400: {"error":"unable to parse 'hank.gpu ': invalid field format"}
Ok, so it seems that the json file for grafana has the localhost "hardcoded" for all attributes. I've just made a merge request to fix that: nicolargo/glances#1684
hi every one!
I'm young student in computer sciences from Libreville(Gabon Africa). I have some class exercice where i have obligation to code(implement) a simple(mini) application monitoring using C or python but still now i can't found some tutorial to do my class exercice. Please some one can help me.
Thanks a lot.
Hi Every1, I got glances to do a restful post to my nodered http-in node but glances still runs the cli ui and keeps posting to the service in short cycles. I tried to find a solution in the documentation but no information about how to run glances, do the POST once and then shuts down. Any ideas, what did I miss?
Hi @all,
we´re developing a mobile app interacting with the glances-API.
Have a look at: https://glutter.de or https://github.com/glutter-dev-team/glutter
we´re developing a mobile app interacting with the glances-API.
Have a look at: https://glutter.de or https://github.com/glutter-dev-team/glutter
I've created new display for the terminal window display. I'm trying to get this display to appear as a web page. I'm cannot figure out how to do this. I get the default view all the time. What do I need to do to get this new view into a web page?
We have a nice little solution which uses the restful export
Does anyone know how we could also test whether mysql (or any other DB) ius running and run a query on it, and include that within the reuslts
or query systemctl status mysql
?
@matyhaty
Have you checked the documentation at https://glances.readthedocs.io/en/latest/aoa/amps.html
Looks like you can define
Have you checked the documentation at https://glances.readthedocs.io/en/latest/aoa/amps.html
Looks like you can define
commands there, that will be run by glanes
thank you :) btw i have a raspberry pi with ds18b20 sensors. when i ran glances directly on the pi it was able to get the sensor's information.. however, now its running on docker and they are not visible. how can i get ds18b20 1-wire sensors ve visible in glances through a docker enviornent?
thank you :)
does anyone know how to secure with user and password glance installed in console mode with: $ docker run -d --name=Glances --restart="always" --network=host -e GLANCES_OPT="-w" -v glances_config:/glances/conf -v /var/run/docker.sock:/var/run/docker.sock:ro --pid host docker.io/nicolargo/glances
Hi, I am using glances on a 32 core dual cpu system, and the sensors plugin has multiple sensors with the same name, and when one sensor reaches a temperature limit and highlites, all of the other sensors with the same name highlite as well, even though they are below the defined temperature
i am using glances v3.1.5 with psutil v5.7.2 on arch linux
curl -L https://bit.ly/glances | /bin/bash
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 167 100 167 0 0 82 0 0:00:02 0:00:02 --:--:-- 82
100 5321 100 5321 0 0 1395 0 0:00:03 0:00:03 --:--:-- 5191
Detected system: Ubuntu
sudo apt-get install -y python-pip python-dev python-docker gcc lm-sensors wireless-tools
Reading package lists... Done
Building dependency tree
Reading state information... Done
Note, selecting 'python-dev-is-python2' instead of 'python-dev'
E: Unable to locate package python-pip
E: Unable to locate package python-docker
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 167 100 167 0 0 82 0 0:00:02 0:00:02 --:--:-- 82
100 5321 100 5321 0 0 1395 0 0:00:03 0:00:03 --:--:-- 5191
Detected system: Ubuntu
sudo apt-get install -y python-pip python-dev python-docker gcc lm-sensors wireless-tools
Reading package lists... Done
Building dependency tree
Reading state information... Done
Note, selecting 'python-dev-is-python2' instead of 'python-dev'
E: Unable to locate package python-pip
E: Unable to locate package python-docker
Has anyone experienced a certain behavior, when running in browser mode, that when selecting a server, the cursor just scrolls to the bottom of the page, jumps back to the top, scrolls to the bottom, etc. No server data is actually shown. All servers have glances 3.1.5 installed, and are running on a mix of Debian 9 and 10. Glances on all were install via pip.