Update 6 BETA - Grafana Reporting Is Finally Here!

[N_G]

If you read the blog post it will say very clearly that you can use self hosted / installed Postgres

 

[SweetAction]

Yes no problem. Anyway, in today's ecosystem, what is the difference between cloud and non cloud? They are both "just another nonbranded machine instance"

So if you have an instance on amazon running postgres and another machine in a private area of your infra, these are still the same technology to offload because the underlying database is postgres. We know that there are many machines hosted with Databases on them that are not necessarily on Google, AWS or Azure. Next step is to put support for all these via the Postgres Plugin.
We thought of this already. Yes, We will definitely add support for offload to SQL on any machine you want. There is no difference for us if it is on premise or cloud.

Agreed, a SQL DB engine is the same engine regardless of in your datacenter or in Amazons, thanks for the clarity regarding support.

You can already use an on prem Postgres no problem at all. Existing or new. Instructions are for new but you can use existing. Definitely we will make support for on prem MS SQL too.

Thanks for the info!

 

[gpaillot]

Yes no problem. Anyway, in today's ecosystem, what is the difference between cloud and non cloud? They are both "just another nonbranded machine instance"

So if you have an instance on amazon running postgres and another machine in a private area of your infra, these are still the same technology to offload because the underlying database is postgres. We know that there are many machines hosted with Databases on them that are not necessarily on Google, AWS or Azure. Next step is to put support for all these via the Postgres Plugin.
We thought of this already. Yes, We will definitely add support for offload to SQL on any machine you want. There is no difference for us if it is on premise or cloud.

I am guessing, installing Grafana in parallel (as in, on the same machine) to 3CX and 3CX's postgreSQL is considered a bad idea ?

 

[UCMUserAZ]

You can already use an on prem Postgres no problem at all. Existing or new. Instructions are for new but you can use existing. Definitely we will make support for on prem MS SQL too.

Thank you on this. I see. Can just create the on-prem postgres and bypass the instructions on the AWS db setup. As long as 3cx has the information to connect. I will test this out too. Thanks!

 

[UCMUserAZ]

All worked with PostGres!. Thanks.

 

[UCMUserAZ]

This would be a nice option for Enterprise licenses to use on-prem SQL. I am still going to setup the BigQuery, but larger orgs may like this option. Grafana is very powerful and was pretty easy to setup. These types of updates are great where you are not having to do full/complete re-installs or uplifts like legacy vendors have you do.

PostGres worked just as well on-prem. Thanks.

 

[SweetAction]

I am guessing, installing Grafana in parallel (as in, on the same machine) to 3CX and 3CX's postgreSQL is considered a bad idea ?

Yes, a bad idea.

 

[Tony_Lin]

This is great! Is there a version in English? I imported this one and it worked great but much of it is in Chinese.

Try this.

 

[Tony_Lin]

The Data Connectors of 3CX have excellent support for PostgreSQL.
We tested serverless databases (such as Supabase, Neon, etc.), and they work properly. The free plan is sufficient for our testing and demonstration purposes.

 

[Michael-CWS]

Hi All ,

To start of, I really like the grafana solution together with the google BigQuery.
Very Nice !

I have a question regarding the metrics / value definitions. In the Dashboard example templates I see "% Missed Call" and then "Abandoned". In the past actual call attempts or pick-up opportunities were counted, rather then an actual "lost call", which meant that a single call could produce "multiple abonded calls". We all know the discussion.
Is there now a possibility to report on actual lost calls and as such have a reachabilty metric that counts only actual calls and not call attempts.
It would save me a lot of headache if I can report on the actual reachability instead of having discussions on what is an abandoned call with management.

Thank you,

 

[Vinnied]

Azure Data Lake is something we are researching - yes for use with Powerbi. Problem here is that from our research,PowerBI direct to Azure data lake will be slow for large storage files and little bit limited for what we intend to achieve.
So then we would need to go the PowerBI via Synapse to ADL and this is an extra complication. For now Azure will be positioned last.

Re BigQuery - Yes you should expect a Json File. Yes it definitely requires Google Workspace.You should consider doing this because it is a great advantage over traditional Google services and will definitely help you.

If you want to use the google stuff, then go in Google Bigquery and tell it to scape the data from your postgres instance.. This is all new stuff still in preview.

Basically it's "nearly" exactly what we did, but the reverse. So instead of 3CX Pushing to Bigquery, Bigquery pulls from Postgres. The drawback is that you would need to open the port from postgres. But if you want to try, this is your best shot until you get Google workspace. Then you can offload the data yourself.

View attachment 47536

Azure Data Lake would be a huge + so we can make reports with Power Bi.
One of custumers is already working in Power Automate

 

[orlanIMP]

Hi everyone,
It looks like we’ll be able to use a Linux server with PostgreSQL instead of relying on AWS.
However, I think I might be missing something.
PostgreSQL works similarly to MySQL or MariaDB — we need to create a database, define tables, specify columns and data types, etc.
So my question is:
Do you have a "template" or example schema for storing CDR (Call Detail Record) files?
Without that, I’m not sure how the server would know where or how to insert the CDR data into the database.

Thanks in advance for your help!

 

[Alphabetic]

Hi everyone,
It looks like we’ll be able to use a Linux server with PostgreSQL instead of relying on AWS.
However, I think I might be missing something.
PostgreSQL works similarly to MySQL or MariaDB — we need to create a database, define tables, specify columns and data types, etc.
So my question is:
Do you have a "template" or example schema for storing CDR (Call Detail Record) files?
Without that, I’m not sure how the server would know where or how to insert the CDR data into the database.

Thanks in advance for your help!

I think they mentioned bringing out some documentation that would help facilitate this

 

[Evolute IT]

Hi everyone,
It looks like we’ll be able to use a Linux server with PostgreSQL instead of relying on AWS.
However, I think I might be missing something.
PostgreSQL works similarly to MySQL or MariaDB — we need to create a database, define tables, specify columns and data types, etc.
So my question is:
Do you have a "template" or example schema for storing CDR (Call Detail Record) files?
Without that, I’m not sure how the server would know where or how to insert the CDR data into the database.

Thanks in advance for your help!

It creates the tables automatically.

 

[nikolascx]

This is great! Is there a version in English? I imported this one and it worked great but much of it is in Chinese.

Set of postgress coming up soon.
Check this overview dashboard - import it and let me know
We will have the others very shortly..
this is for PostgeSQL
Put the Table path without the IP and port in the top of the dashboard.

 

[nikolascx]

So, there is no way to self-host this ? we have to use (paid) cloud services ?

What prices to expect for the AWS db server ?

You can self host it yes.

AWS database pricing depends on several factors, and it can get quite complex. Here are the key things that affect cost:

1. Database engine – (e.g. PostgreSQL, MySQL, Aurora, etc.)
2. Instance type – More powerful servers (CPU, RAM) cost more.
3. Storage type & size – SSD vs magnetic, provisioned IOPS, etc.
4. Region – Prices vary by AWS region.
5. High availability – Enabling replicas or multi-AZ increases cost.
6. Usage model – On-demand vs reserved vs spot pricing.

For a basic idea:
A small PostgreSQL instance (like db.t3.micro) might cost $10–15/month.
A production-grade instance with replication and storage could be $100–$1000+ per month.

I recommend using the world renowned AWS Pricing Calculator for an accurate estimate based on your specific needs.
https://calculator.aws/#/

 

[nikolascx]

Hi Team

The installation of postgreSQL creates everything for you.
Then you configure.

Now. thanks to the update we made in update 5, where we got 8 legacy fat tables and flattened them to 1 nice 40 feature columned table, this makes everything easier when the offloader dumps data from 3CX to your remote postgresql server.

Plus there is a script that you can run on debian and it will do everything for you. So this script mimics the windows installer and asks you questions to get you going.
The answers you give to the script, are the same answers you have to give when you configure the Grafana PostgreSQL Plugin.

We are working on some training. I will make sure to include a demo of this.
Also let me know your questions

 

[nikolascx]

I am guessing, installing Grafana in parallel (as in, on the same machine) to 3CX and 3CX's postgreSQL is considered a bad idea ?

Why not yes - If it is only for you, its great..
Is it a busy system? It depends on what load you process and what resources you have available.

 

[nikolascx]

To all Postgress users that want to get started on a remote postgreSQL Machine

Postgress youtube + script to prepare a debian machine with Postgres installation.

Get script from here
https://github.com/develops20/debian-postgresql-setup-script

All this is beta but good to get started.

Working on Grafana dashboards on Postgres. Stay tuned

 

[UCMUserAZ]

Hi everyone,
It looks like we’ll be able to use a Linux server with PostgreSQL instead of relying on AWS.
However, I think I might be missing something.
PostgreSQL works similarly to MySQL or MariaDB — we need to create a database, define tables, specify columns and data types, etc.
So my question is:
Do you have a "template" or example schema for storing CDR (Call Detail Record) files?
Without that, I’m not sure how the server would know where or how to insert the CDR data into the database.

Thanks in advance for your help!

This was easy. 3cx creates the tables, etc once you setup the data connector to it from within 3cx.