Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 07:57
    nicolargo assigned #1985
  • Jan 28 15:54
    nicolargo labeled #2198
  • Jan 28 15:47
    nicolargo closed #2249
  • Jan 28 15:47
    nicolargo milestoned #2248
  • Jan 28 15:47
    nicolargo milestoned #2248
  • Jan 28 15:47

    nicolargo on develop

    Update __main__.py non-executa… Merge pull request #2248 from a… (compare)

  • Jan 28 15:47
    nicolargo closed #2248
  • Jan 28 15:45
    nicolargo closed #2241
  • Jan 28 15:45
    nicolargo locked #2241
  • Jan 25 09:42

    renovate[bot] on webpack-cli-5.x

    (compare)

  • Jan 25 09:41
    renovate[bot] closed #2200
  • Jan 25 09:41
    renovate[bot] edited #2200
  • Jan 25 09:41
    renovate[bot] edited #2163
  • Jan 25 08:10
    nicolargo edited #2252
  • Jan 25 07:29
    nicolargo milestoned #2255
  • Jan 25 07:29
    nicolargo milestoned #2255
  • Jan 25 07:29

    nicolargo on develop

    moved now plugin after sensors … bump dev dependencies bump dependencies and 1 more (compare)

  • Jan 25 07:29
    nicolargo closed #2157
  • Jan 25 07:29
    nicolargo closed #2255
  • Jan 25 07:28
    nicolargo closed #2256
Matt Hiscock
@matyhaty
(last message was a month ago!)
steschuser
@steschuser
who dares to disturb us?
Matt Hiscock
@matyhaty
Haha, Well now thatyour disturbed...

I seem to be struggling with Glances a little. Im looking to be super selective about what I would like Glances to send to (for example influx, but would rather it hit a api endpoint on another server).
I would like only for it to push out:

  • CPU
  • Load (x,x,x)
  • Ram
  • Perhaps something on mysql or postgres
  • Diskspace

Is that kinda thing even possible?

Why...
I want to create a dashboard of servers (we have well over 40) which are just simple panels, one for each - which can just show basic info, and alert us to any problem. Grafana is nice, but way too in depth. If we see a problem we will login to the server/droplet and investigate from there.

I know its annoying to get long messages, but if anyone can point me in the direction, ill go from there!

steschuser
@steschuser
Dont think its possible to limit the data glances sends.
maybe best to use curl if you only need a few values
Gordon Hsu
@sosonian
Hi, every one , I am new to Glance. And I have a quesiton, what is the difference between the toal cpu usage and cpu usage in processes list?
Gordon Hsu
@sosonian
CPU usage of PID9303 is 100%, but in top left corner, the total cpu usage shows only 3.9%,why?
Akshay Verma
@aeroaks
How do I add export configuration of CSV in the '.conf' file?
sshaikh
@sshaikh
I'm installing glances and following the instructions in the docs (and bash script)
with things like pip install 'glances[action,browser,cloud... ]' is there a list of libraries available and descriptions of what they do?
sshaikh
@sshaikh
Another question - can the webserver be set as a client browser?
aadewojo
@aadewojo
Hi All, I am new here and new to glances

I want to use it to monitor some virtual machines, however I am not sure it works except the standalone. I tried to do the client/server mode operation.

I ran glances -s on a virtual machine I want to use as server and then glances -c @serverip on client machines. After running the command on server, it returns a messages saying Glances XML-RPC server is running on 0.0.0.0:61209
Announce the Glances server on the LAN (using x.x.x.x ip address)
After running the command on the client, it doesn't say anything.

How do I know it is working. How do I get stats of results?
Tia

aadewojo
@aadewojo
@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?
steschuser
@steschuser
@aadewojo
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
@aadewojo
@steschuser , thank you. I will work on it
nashid
@nashid
I installed glaces the easy way i.e. brew install glances. However I need to monitor gpu usage. Whats the option for that?
aadewojo
@aadewojo
@steschuser , hi, please is influxdb same as influxdb cloud?
aadewojo
@aadewojo
@steschuser , also did you have to install influxdb on all the machines each?
steschuser
@steschuser
@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
aadewojo
@aadewojo
@steschuser , thank you very much for the link. I managed to install all.
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?
Ben Doerr
@bendoerr
Has anyone tried using pyinstaller or cx_freeze to create a standalone bin? I'd love to use the glances API but embedded linux without much storage, trying to avoid installing python.
Paulo Flabiano Smorigo
@pfsmorigo_gitlab
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.
Paulo Flabiano Smorigo
@pfsmorigo_gitlab
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"}
Paulo Flabiano Smorigo
@pfsmorigo_gitlab
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
aadewojo
@aadewojo
Hi All, please I need some help with using glances to monitor many virtual machines. Where do I specify the virtual machine's details? DO I have to run glances on each of the virtual machine? Thanks
gandhy241
@gandhy241

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.

JeromeDevome
@JeromeDevome
Hello, thanks for the work. Can we keep glances web open?
stroehmi
@danielstroehmann
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?
yairmiz
@yairmiz
Hi Everyone, I'm running the latest glances 3.1.5, having a wierd issue with exporting container data to influxdb, I can see in the client the containers cpu and memory consumption but for some reason I dont get these values exported to influxdb, only the container names
saitoh183
@saitoh183
hi
saitoh183
@saitoh183
what is the proper syntax for connecting glances to influxdb over SSL on a remote server?
jh
@realjax
Hello, quick question.. can I also build/install this without the cassandra driver ?? Would save me an hour of waiting :(
jh
@realjax
nm, found it :)
Moritz
@MrzJkl
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
rltheis27
@rltheis27
Hello everyone,
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?
steschuser
@steschuser
@MrzJkl noice
Matt Hiscock
@matyhaty
Hi Glancers
We have a nice little solution which uses the restful export