Want #cloudnative but with the power of #declarative configuration? The recoverability of #transactions for system configuration?
Wednesday it's the online #guix meet-up! With a great talk by @paulbutgold
about running docker / oci containers using the Guix configuration system.
His Gocix project has #prometheus, #grafana, #forgejo, #conduit and #traefik examples.
Meet-up details:
Some self-hostable alternatives for well-known cloud services, like #jira or #confluence.
#selfhosted #plane #docmost #n8n #strapi #grafana #nocodb #projectmanagement #notes #wiki #automation #cms #analysis #datastorage
#alternative
Kennt ihr zufällig jemensch oder Firma, die gegen Geld Dienstleitungen rund um #Grafana anbietet?
Wir - aka. die Firma - sucht Hilfe bei der Erstellung von Grafana Notification Templates.
If you're coming to #FOSDEM and like #3dprinting, make sure to stop by the #Prusa booth (next to the #Grafana booth), in building H this time.
I have it on good authority that they'll show off the Prusa #CoreOne this time.
Hostet hier ein wunderbares Wesen einen #APNS Server, den ich für meinen #ntfy und mein #Uptimekuma bzw. den #Grafana benutzen kann?
Ich hab keinen Apple Dev Account, kann mir also keinen selber hosten...
I've just finished a #Grafana #vps #monitoring panel for our Webserver.
I've created this explanatory panel for the telemetry...I've started with a "Hamster wheel" for the CPU...
...and then kinda got carried away with the analogy
#uptime
Argh... Working with #grafana V 11.3 is so frustrating.
1. Once you figure out all the panels you use are linked. So changing the query on one, changes them all on the others. First gotcha [You have to unlink the panel]
2. Then you encounter some wacky, buggy behaviour editing panels.
Changing query and parameters will fail to save the panel and you have to do some silly shenanigans like go to view the JSON file, or so something "meaningful" that the code then decides to save the panel.
3. And then there is a failure mode when the query changes take effect...but not the graph graphic attributes.
The workaround I found is to duplicate the panel, and the duplicate will properly execute the change.
Surely I am not the only one frustrated so?
I'm running a Docker stack on Debian.
Got my hands dirty with #CheckPoint #skyline Monitoring today. Skyline monitors Check Point security devices using #opentelemetry with #Prometheus and #Grafana. An OpenTelemetry CPView Agent service runs on the security devices to collect and export the health metrics.
You can find more details and links on the secure knowledge base article: https://support.checkpoint.com/results/sk/sk178566
What do you think?
I'm getting pretty close to my perfect #grafana server monitor panel.
Because I'm running a #Docker stack
There is a breakdown of performance by container.
The CPU breakdown is nice too...
Especially for a #VPS where the STEAL and IOWAIT are important because you don't want them above zero as it means the #hypervisor is struggling/overloaded.
I would like to add a list of top 3-5 processes, like you get on with "top".
But it's not straight forward.
The idea is you have all the servers under your command cycling in a playlist.
N.B. TIL that the entropy on a VPS is only around 200 (1000 recommended) which means #encryption on a VPS sucks.
Going to look into implications for #ssh and such later.
#uptime #monitoring #prometheus
@schenklklopfer
(You may need to download the image to see it in detail)
I think I may have a solution for you.
You can set up a Playlist of multiple Dashboards in #grafana
You can also feed data from other servers to your grafana server...allegedly. I have not done it yet, because I'm tweaking my perfect server monitoring panel.
So you can (very likely) have multiple server dashboards flicking through your playlist. Or have multiple panels for multiple serves in a row for each server.
e.g.
DASHBOARD
CPU_SRV1 DISK_SRV1 NET_SRV1
CPU_SRV2 DISK_SRV2 NET_SRV2
CPU_SRV3 DISK_SRV3 NET_SRV3
#telegraf or #cadvisor
To monitor #docker stacks?
#portainer is nice for standalone, but if you want to plug into #grafana?
The problem with #Grafana on mobile devices is the screen resolution.
Even a basic dashboard gets squished to fuckery.
You probably could squeeze a bit of screen real estate by playing with margins. But the main problem is, you can't easily fix the dashboard display size, it always scales, and I tried a couple of methods.
TLDR; No
Ok...so you know that in #grafana you can not directly change the Dashboard JSON descriptors by directly editing them, because FSCK YOU!?
And finding attributes on the right panel to tweak can be painful for novice/casual users? And sometimes buggy too?
WORKAROUND: Export the whole Dashboard to external use...Edit the MOFO, then import it as a new Dashboard.
GG!
#uptime #monitoring
#Grafana picking the least contrasting colors for the the most important lines on your graphs:
Gibt es nach wie vor keine #Grafana App für #iOS?
Es gibt etwas, das nennt sich "Grafana IRM", ist wohl nur fürs Alerting - besser als nix, aber geht scheinbar nicht mit der kostenlosen Version, weil es diesen "IRM" Tab, den es geben sollte nicht gibt...
Sooo...?
Wie mant eins das dann von unterwegs?
Die Webseite ist nicht gut genug aufm Handy...
With support for #postgresql 17, #ansible 2.17, an upgraded #monitoring stack, and even support for a whole new #linux #distribution, you really should check out my write-up on what's new in Crunchy #Postgres via Automation V2.2 https://hunleyd.github.io/posts/crunchy-postgres-via-automation-v2.2/
If you are developing content for a Grafana instance, you may have experienced this. I have faced this problem a few times.
When I develop a Grafana plugin on my local development machine, I like to have representative telemetry data.
Connecting to production data sources may be impossible, but there are other ways to connect.
You can use Grafana as proxy, see
https://ceesbos.nl/posts/20241108-grafana-as-proxy/ to find out how you can do that.
I like to monitor all sorts of metrics with #prometheus and #grafana. How are people monitoring their #freebsd #bhyve vms and #jails? There doesn’t seem to be a Prometheus exporter for those. Do I need to make one? My Google-fu seems to be failing here.
It's the season of promotions at #Grafana Labs, and I'm very happy to see two engineers promoted, from senior to staff and from staff to principal respectively, purely on their upstream open source work on #Prometheus
There are not a lot of companies left which not only put money where their mouth is and sponsor upstream work but also promote people internally for external work.
