# InfluxDB UI for Visualization
###### tags: `Independent Study`
:::success
**Reference:**
[Influxdata Document - Chronograf](https://docs.influxdata.com/chronograf/v1.8/introduction/getting-started/)
[Docker Hub - chronograf](https://hub.docker.com/_/chronograf/)
[Influxdata Document - Basic C++ client](https://w2.influxdata.com/blog/getting-started-c-influxdb/)
[Influxdata Document - C++ Client github](https://github.com/offa/influxdb-cxx)
[Influxdata Document - Client library](https://docs.influxdata.com/influxdb/v1.8/tools/api_client_libraries/)
[O-RAN Virtual Exhibition](https://www.virtualexhibition.o-ran.org/classic/generation/2021/category/intelligent-ran-control-demonstrations/sub/intelligent-control/118)
[influxdb v1 - issue](https://github.com/influxdata/influxdb/issues/11035)
[influxdb1-clientv2 github](https://github.com/influxdata/influxdb1-client)
**Note:**
We are using the old version of InfluxDB, so we have to install the old UI (Chronograf), open the UI interface, and configure the dashboard to make sure the data can be displayed on the dashboard.
In this page, I will
1. Introduce why InfluxDB.
2. Roughly explain the data I wanna show about independant project.
3. Show how could data can be visualized on chronograf.
4. Provide the configuration and source code that xApps can access InfluxDB.
:::
[TOC]
## InfluxDB Introduction
1. InfluxDB is a time-series database used to store runtime data which means data would change over time, e.g., rsrp, rsrq, throughput..
2. The database used in cherry release is Redis used to store key-value data which means data wouldn't change over time, e.g., the name of gnb, plmn ID...
3. Combining the first and second points, you can see why we need to use InfluxDB because Near-RT RIC wants to analyze the data that changes in time, and then make adjustments of operation in RAN.
4. By the way, compared with PromQL, InfluxDB has UI for monitoring visualizations, and most importantly, it has a durable long-term storage.
## What I want on influxDB
I will explain what/why I wanna show on InfluxDB this part.
### Cell Load with Slice - Gauge
To make the manager can visually know the how RIC adjust cell load while the load balancing algorithm is implementing.
The pictures below roughly shows what I wanna show on InfluxDB. Gauge shows current cell load and slice load. Graph shows cell load with timestamp, cell load increment, and offloaded cell load. Currently, I don't have data about slice so I am not exactly know how to present it. Thus, it will be put aside.
### Cell Future Throughput UE belongs to - Graph
To make the manager can visually know the relation between UE downlink throughput and other UE Data, e.g., signal strength, signal to noise ratio..
The picture below show what I wanna show on InfluxDB.
## Create a InfluxDB UI
The old version of InfluxDB(1.8.0) does not have a UI, so you have to install Chronograf and connect to it. The new version is already integrated, but OSC seems to use the old version of InfluxDB, so I installed Chronograf.
### Step 1 - Install Chronograf
```bash=
docker pull chronograf:1.9
# docker pull grafana/grafana
```
### Step 2 - Check the IP & Port of InfluxDB
```bash=
kubectl get service -n ricplt
```
### Step 3 - Run Chronograf
```bash=
docker run -p 8888:8888 --net=influxdb chronograf --influxdb-url=http://10.97.7.71:8086
# docker run -d --name=grafana -p 3000:3000 grafana/grafana
```
### STEP 4: Run AD xApp
### <Center>[Installation Guide (E-Release) ](https://hackmd.io/@OP6n8eNbREK5elXuZw1wBw/Hy1RaOFTK)<Center>
### Outcome
The outcome below shows the data is indeed linked to the database.
## Influxdb Client - C++
Since I might use TS xApp to upload the cell load, we need to use the C++ library of InfluxDB. Below shows the installation step to use it, and the basic writing data method to the InfluxDB by using c++.
### Step 1 - Install cmake
```bash=
sudo apt-get update
sudo apt-get install libssl-dev
sudo apt-get install -y libcurl4-openssl-dev
sudo apt-get install -y autoconf gawk libtool automake pkg-config autoconf- archive
sudo apt install openssl
sudo apt install libsctp-dev
cd
wget https://github.com/Kitware/CMake/releases/download/v3.16.2/cmake-3.16.2.tar.gz
tar -xzvf cmake-3.16.2.tar.gz
cd cmake-3.16.2.tar/
./bootstrap
make
sudo make install
cd
cp ./cmake-3.16.2/bin/cmake /usr/bin/
```
### Step 2 - Install InfluxDB Package
location: ts/Dockerfile
```bash=
apt-get update
apt install -y curl
apt-get install -y libcurl4-openssl-dev
apt-get install libboost-all-dev
cd
git clone https://github.com/offa/influxdb-cxx.git
cd influxdb-cxx
mkdir build && cd build
cmake -D INFLUXCXX_TESTING:BOOL=OFF ..
sudo make install
```
### Step 3 - Modify Dockerfile
location: ts/Dockerfile
```dockerfile=
# # snarf up SDL dependencies, then pull SDL package and install
# RUN apt-get update
# RUN apt-get install -y libboost-filesystem1.65.1 libboost-system1.65.1 libhiredis0.13
# RUN wget -nv --content-disposition ${PC_STG_URL}/sdl_${SDL_VER}-1_amd64.deb/download.deb && \
# wget -nv --content-disposition ${PC_STG_URL}/sdl-dev_${SDL_VER}-1_amd64.deb/download.deb &&\
# dpkg -i sdl-dev_${SDL_VER}-1_amd64.deb sdl_${SDL_VER}-1_amd64.deb
RUN git clone https://github.com/Tencent/rapidjson && \
cd rapidjson && \
mkdir build && \
cd build && \
cmake -DCMAKE_INSTALL_PREFIX=/usr/local .. && \
make install && \
cd ${STAGE_DIR} && \
rm -rf rapidjson
######### Ken Install influxDB ###
RUN apt-get update
RUN apt install -y curl
RUN apt-get install -y libcurl4-openssl-dev
RUN apt-get install -y libboost-all-dev
RUN git clone https://github.com/offa/influxdb-cxx.git
RUN cd influxdb-cxx && mkdir build && cd build && cmake -D INFLUXCXX_TESTING:BOOL=OFF .. && make install
RUN ldconfig
# install curl and gRPC dependencies
```
### Step 4 - Modify CMakeList.txt
location: ts/src/ts_xapp/CMakeList.txt
```cmake=
find_package(Protobuf REQUIRED)
add_executable( ts_xapp ts_xapp.cpp )
######### Ken Install influxDB ###
link_libraries(InfluxDB)
find_package(InfluxDB)
target_link_libraries( ts_xapp InfluxDB;ricxfcpp;rmr_si;pthread;curl;rc-api;grpc++;${Protobuf_LIBRARY})
install(
TARGETS ts_xapp
DESTINATION ${install_bin}
)
```
### Step 5 - Copy InfluxDB Header Files
```bash=
cd
cp -a /usr/local/include/ ts/src/ts_xapp/
```
### Step 6 - Modify Source Code for testing (Write Data)
```cpp=
#include <grpcpp/security/credentials.h>
#include "../../ext/protobuf/api.grpc.pb.h"
//----- Ken Create the Route for InfluxDB
#include "InfluxDBFactory.h"
using namespace rapidjson;
using namespace std;
using namespace xapp;
using Namespace = std::string;
using Key = std::string;
using Data = std::vector<uint8_t>;
using DataMap = std::map<Key, Data>;
using Keys = std::set<Key>;
//----- Ken Create the Route for InfluxDB
string influxdb_url = "http://10.97.7.71:8086?db=UEData";
auto db_influx = influxdb::InfluxDBFactory::Get(influxdb_url);
```
```cpp=
extern int main( int argc, char** argv ) {
//----- Ken Create the Route for InfluxDB
db_influx->write(influxdb::Point{"test"}
.addField("value", 10)
.addTag("host", "localhost")
);
//-----------------------------------------
int nthreads = 1;
char* port = (char *) "4560";
shared_ptr<grpc::Channel> channel;
```
### Outcome
For the basic operation, as what source code express, we can specify the value, fields, and the tags, uploading the data to the TimeSeries DataBase, InfluxDB, by using c++.
## Influxdb Client - Go
Since I might use KPIMON xApp to upload the data from RIC Test, we need to use go library of InfluxDB. Below shows the installation step to use it, and the basic writing data method to the InfluxDB by using Go.
### Step 1 - Modify the Dockerfile
location: kpimon/Dockerfile
```dockerfile=
# "COMPILING E2SM Wrapper"
RUN cd e2sm && \
gcc -c -fPIC -Iheaders/ lib/*.c wrapper.c && \
gcc *.o -shared -o libe2smwrapper.so && \
cp libe2smwrapper.so /usr/local/lib/ && \
mkdir /usr/local/include/e2sm && \
cp wrapper.h headers/*.h /usr/local/include/e2sm && \
ldconfig
WORKDIR /go/src/gerrit.o-ran-sc.org/r/scp/ric-app/kpimon
RUN mkdir pkg
RUN go env -w GO111MODULE=off
###
######### Ken Install influxDB ###
RUN go get -u github.com/influxdata/influxdb1-client/v2
####
RUN go build ./cmd/kpimon.go && pwd && ls -lat
FROM ubuntu:20.04
```
### Step 2 - Modify the source code for testing (Write Data)
location: kpimon/control/control.go
```go=
import (
"encoding/json"
"errors"
"log"
"os"
"strconv"
"strings"
"sync"
"time"
"gerrit.o-ran-sc.org/r/ric-plt/sdlgo"
"gerrit.o-ran-sc.org/r/ric-plt/xapp-frame/pkg/xapp"
//----- Ken Create the Route for InfluxDB
"github.com/influxdata/influxdb1-client/v2"
"fmt"
//"github.com/go-redis/redis"
)
```
```go=
func (c *Control) sendRicSubRequest(subID int, requestSN int, funcID int) (err error) {
//----- Ken Create the Route for InfluxDB
cl, err := client.NewHTTPClient(client.HTTPConfig{
Addr: "http://10.97.7.71:8086",
})
if err != nil {
fmt.Println("Error creating InfluxDB Client: ", err.Error())
}
defer cl.Close()
// Create a new point batch
bp, _ := client.NewBatchPoints(client.BatchPointsConfig{
Database: "UEData",
})
// Create a point and add to batch
tags := map[string]string{"kpi": "kpi-total"}
fields := map[string]interface{}{
"idle": 10.1,
"system": 53.3,
"user": 46.6,
}
pt, err := client.NewPoint("cell_usage", tags, fields, time.Now())
if err != nil {
fmt.Println("Error: ", err.Error())
}
bp.AddPoint(pt)
// Write the batch
cl.Write(bp)
//------------------------------------------------------------
var e2ap *E2ap
var e2sm *E2sm
```
### Outcome
For the basic operation, as what source code express, we can specify the value, fields, and the tags, uploading the data to the TimeSeries DataBase, InfluxDB, by using go.
## Influxdb Client - Python
In AD & QP xApp, they can already access the InfluxDB by using python, so I will just list the setting and source code to access InfluxDB by using AD source code as an example.
### Step 1 - Modify the setting
location: setup.py
```python=
from setuptools import setup, find_packages
setup(
name="test",
'''
Some settings here..
'''
install_requires=["A","B","C", "influxdb"],
'''
Some settings here..
'''
)
```
### Step 2 - Import InfluxDB Package
```python=
from influxdb import DataFrameClient
import pandas
```
### Create DataBase
```python=
class INSERTDATA:
def __init__(self):
host = 'r4-influxdb.ricplt'
self.client = DataFrameClient(host, '8086', 'root', 'root')
self.dropdb('UEData')
self.createdb('UEData')
def createdb(self, dbname):
print("Create database: " + dbname)
self.client.create_database(dbname)
self.client.switch_database(dbname)
def dropdb(self, dbname):
print("DROP database: " + dbname)
self.client.drop_database(dbname)
def dropmeas(self, measname):
print("DROP MEASUREMENT: " + measname)
self.client.query('DROP MEASUREMENT '+measname)
```
### Write & Cluster(Tag) Data
```python=
db.client.write_points(data, 'valid')
#####################################################################
db.client.write_points(df, 'train', batch_size=500, protocol='line')
#####################################################################
db.client.write_points(df, 'liveUE', batch_size=500, protocol='line')
#####################################################################
def write_anomaly(self, df, meas='AD'):
"""Write data method for a given measurement
Parameters
----------
meas: str (default='AD')
"""
self.client.write_points(df, meas)
#####################################################################
```
### Read Data
```python=
result = self.client.query('select * from ' + meas + ' limit ' + str(limit))
```
### Outcome
## Chronograf
### Template Variable
[Ref](https://docs.influxdata.com/chronograf/v1.9/guides/dashboard-template-variables/#quoting-template-variables-in-influxql)
```
# at the dashboard , import csv
# Name = ueid
# at the Panel
SELECT mean("nr_cell_slice_usage_0_embb") AS "mean_nr_cell_slice_usage_0_embb" FROM "UEData"."autogen"."slice_usage" WHERE "ueid" = ':ueid:' and time > :dashboardTime: AND time < :upperDashboardTime: GROUP BY time(:interval:) FILL(null)
```
## Grafana
### Template Variable
[Ref](https://grafana.com/docs/grafana/latest/variables/variable-types/chained-variables/)
```
# at the setting, use type "Query"
SELECT ("ueid") FROM "UEData"."autogen"."slice_usage"
# at the Panel
WHERE ueid =~ /^$ueid$/
```
### Show Prediction
```
```
## Add Influxdb source code into the project (cmake)
**增加一個專案到一個專案** (那個專案要有寫好的CMakeList.txt)
```
add_sub_directories()
```
Sign in with Wallet
Connect another wallet