Novu on Grafana
Is there any way to scrape the metrics of novu on grafana for monitoring of novu?
32 Replies
@Jahnavi, you just advanced to level 1!
I am interested in that as well. Thanks.
Please let me know if you do any progress
@Jahnavi
We currently don't have this feature and I see we are not planning this in upcoming cycle.
Would you like to share why you want to monitor Novu?
Are you self hosting the Novu?
Getting all the metrics on grafana of whatever different sites we are using.we are hosting novu on kubernetes
Thanks for sharing the usecase @Jahnavi
Currently, we don't have and we are not planning Grafana native integration in community self host option.
If you are self hosting Novu in kubernetes based environment, then you must be using Novu on large scale system. I would like to recommend you having a look on commercial self host option with advanced features and better support SLAs
https://docs.novu.co/self-hosting-novu/commerical-self-host
@Zac Clifton I have seen you raising a pr https://github.com/novuhq/novu/pull/3000 on the same issue.Can you guide me through of what exactly you have done and why was the pr closed
GitHub
feat(otel/analytics): Add Open-Telemetry and Promethus Montioring t...
What change does this PR introduce?
This PR adds the initial commit of nestjs-otel to Novu's api, ws, and webhook containers.
This system set up a prometheus metric server on 9464 on the path &...
@Pawan Jain can you take a look at the pr and tell me if there will be anything on that in your future works
Let me talk to the team internally on this, there are currently alot of factors that played into this decision (though I understand it may not seem this way).
I would like to have this support in the community version however, as we have our values to be transparent to the community https://handbook.novu.co/,
currently our thought is that this features would be enterprise only, however I (personally separate from the company) do not fully agree with this thought.
As an architect, I have based many decisions on the fact that the system has support for the grafana stack.
We currently only support new relic monitoring. So if you want to monitor Novu you would have to do it this way.
Also I want to be empathic here as I understand that "If you are self hosting Novu in kubernetes based environment, then you must be using Novu on large scale system." is a generalization that is not always true. As you can run K8 for cost saving, burst scaling, etc especially if you have the knowledge on how to maintain and support it as a small company. Which means you do not run at large scale but do have infra set up to scale to that size.
I would love to hear your thoughts on why you think we should support the grafana stack in the community version! This would also help me make the case as we would have interest from the community.
Thank you very much for reaching out 🙏 and sharing your passion with us.
I am happy to take ownership of this request in the mean time and have marked it as unresolved.
I would like to ask that we have this discussion here so we can be open to the community about it.
First of all, a big thanks! I appreciate your prompt response and your efforts to facilitate internal communication regarding this matter. And about the thoughts on Grafana,it has robust data visualization capabilities that empowers users to create highly interactive and insightful dashboards. Coupled with its active and thriving open-source community, Grafana has emerged as the standard tool for monitoring. As we strive to enhance our monitoring dashboards effectively, our goal is to leverage Grafana to consolidate metrics from all applications, ensuring a professional and comprehensive approach to monitoring and visualization.
My pleasure! The community is what got us here and I want to make sure we serve you and the rest of the community responsibly with open arms and ears.
What parts of the grafana stack do you use? loki, tempo, prometheus, mimir, Pyroscope, grafana, on-call?
Do you use on-prem grafana stack or grafana cloud?
What metrics would would want to see from Novu that are not covered in our activity feed?
In our grafana stack we have incorporated a combination of monitoring tools including loki,tempo and victoriametrics.Our entire stack is deployed on our on-premises infrastructure.Regarding the metrics not currently featured in the activity feed, we would appreciate the inclusion of data points such as server requests, server responses, CPU utilization, and client error counts etc.I will engage in discussions with my team to gain further insights and understanding on what more metrics we would need.Thanks
Thank you, this would be very helpful to see the full picture of what we could provide with prometheus support.
Yeah looking forward to it.Thanks 😁
@Jahnavi, you just advanced to level 2!
Last question, for you if we are good with moving forward on grafana support would you be willing to make a PR?
Yeah would love to!
Do you have any documentation on the the new relic monitoring of novu?
Sort of, basically you would add :
to each container and new relic does the rest.
This does set up custom new relic event for each queue which we have custom dashboards for to help monitor them
Your welcome to start working on it if you want, however I am not able to have an internal descussion till next week.
Yeah do let me know when there is any progress.Keep me posted😀
Still have not put out a request to discuss this week but put it on our biweekly tech meeting we will have next wednesday
Hi,
I am stuck on issue that you might have solved
Regarding kubernetes with ingress-nginx
Check out this section of the docs
https://docs.novu.co/self-hosting-novu/deploy-with-docker#reverse-proxy-load-balancers
Hi @Zac Clifton & @Jahnavi
I have already been through this
Mine is specifically related with kubernetes and ingress path
Can you create a new support issue for this?
Sure
For understanding purpose issue is
If ingress-nginx is configured on dns entry Novu.domain.com
And path /novuweb
It routes to that address but the context_path properties doesn't match the request, it routes to / and gives 404
Example Novu.domain.com/env-config.js
Whereas it should have been
Novu.domain.com/novuweb/env-config.js
@aay, you just advanced to level 1!
Hi @Jahnavi
Are you available for discussion
Hey I am available over the weekend we can discuss then
@Jahnavi can we do it today evening or night
I am not available weekend and it is a little urgent
@Jahnavi
Please schedule a meet time
https://calendly.com/aayushave/novu-k8s
Calendly
Novu K8s - Aayush R
@Jahnavi
Didn't heard back from you
@otatopotato has a pr merged for this, but it isn't working for spring gateway, ingress nginx
Closing this post due to inactivity. Feel free to create a new post if you have any question or you are still facing issue