## Overview
This Helm chart provides Kubernetes deployment configurations for [qryn-cloud](https://github.com/metrico) a polyglot, lighweight, multi-standard observability framework for Logs, Metrics and Traces, designed to be drop-in compatible with Loki, Prometheus, Tempo and Opentelemetry.

## Prerequisites
- Kubernetes 1.19+
- Helm 3.7+

## Get Repository Info

```bash
helm repo add qryn-cloud-helm https://metrico.github.io/qryn-cloud-helm/
helm repo update
```

# Installation

1. Create namespace in the kubernetes cluster you want to use: `kubectl create namespace qryn`
2. Create docker registry secret yaml to pull docker images
```
cat <docker-registry-secret.yaml
apiVersion: v1
kind: Secret
metadata:
name: qryn-dckr
namespace: qryn
type: kubernetes.io/dockerconfigjson
data:
.dockerconfigjson: $(base64 -w 0 <<< $(cat <<'EOF'
{
"auths": {
"your-registry.com": {
"username": "",
"password": ""
}
}
}
EOF
))
EOR
```
3. Make the kubectl secret in the qryn namespace to reach the private docker registry as
https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
```
kubectl apply -f docker-registry-secret.yaml
```
4. Create `values.yaml` to override the needed parameters.
Please take a look to the [configuration](#configuration-options). Some of the parameters are required to override.
5. Install helmchart
```
helm install qryn . -f values.base.yaml -f values.yaml --namespace qryn --create-namespace
```

# First database setup

Before using Qryn the database should be set up.
In order to set up the database you need to create an API call to qryn-ctrl.

1. Forward the qryn-ctrl port
`kubectl port forward 8080:8080 -n qryn `
2. Do the API http call to qryn-ctrl:
```bash
cat <",
"node": "clickhouse1",
"pass": "",
"name": "",
"host": "",
"port": ,
"primary": true,
"debug": true,
"secure": ,
"cloud": ,
"cluster_name": "",
"distributed": ,
"ttl_days": ,
}
]
}

EOF
```

# Data deletion

The data deletion functionality is optional and can be configured using the helmchart. It consists of the modules:

- qryn-reader properly configured
- qryn-worker cronjob
- redis-compatible storage of the deletion tasks and deletion status

## Redis-compatible storage

Any redis-compatible instance can be used to store the tasks and synchronize the workers.
The storage should be persistent. Non-persistent storage may lose planned tasks and status synchronization.
The helmchart contains an optional configurable keydb storage with the following values.yaml configuration:

```yaml
keydb:
# enable or disable a keydb instance
enabled: false
# enable or disable a persistent volume
persistence: true
# liveness / readiness / initContainers specification default for any deloyment
livenessProbe: { }
readinessProbe: { }
initContainers:
enabled: false
# keydb password is mandatory for the helmchart
env:
KEYDB_PASS: XXXXXXX
```

These configurations can be overriden with the user values.yaml file.

## Qryn-worker cronjob:

The qryn delete worker cronjob is present in the helmchart with the following values.yaml
```yaml
deleteWorker:
# enable / disable the cronjob
enabled: false
podAnnotations: []
image:
repository: qxip/qryn-deleter
tag: 1.2.91-beta.65
imagePullPolicy: IfNotPresent
# schedule for the cronjob
schedule: "* * * * *"
nodeSelector: { }
tolerations: [ ]
affinity: { }
resources:
requests:
memory: 100Mi
cpu: 100m
limits:
memory: 100Mi
cpu: 100m
```

The schedule configuration can be changed to run the deletion worker less frequently.

## The qryn-go configuration

The following variables were added to the configuration configmap:
```yaml
# Delete configuration
QRYN_WORKER_TYPE: "redis"
QRYN_WORKER_SYNC_URL: "redis://XXXX@qryn-keydb:6379"
```

The worker type should be "redis".
The sync url should be a valid redis URI.

# Ingress setup
In order to reach Qryn from outside, the http ingress rules should be configured.
Current helm has a default ingress definition. It can be enabled for writer and reader separately by configuring
`reader.ingress` and `writer.ingress` configuration. Please look at the [configuration](#configuration-options).

The http routes should be configured for Qryn reader and writer services are as follows.

For qryn-writer service (regexp notation):
- /loki/api/v1/push
- (/api)?(/v1)?/prom/remote/write
- /tempo/spans
- /tempo/api/push
- /influx/api/v2/write
- /v1/traces
- /[^/]+/_doc(/[^/]+)?
- /[^/]+/_create/[^/]+
- (/[^/]+)?/_bulk
- /api/v2/spans
- /api/v2/logs
- /api/v2/series
- /cf/v1/insert

For qryn-reader service (regexp notation):
- /ready
- /metrics
- /config
- /api/v1/metadata
- /api/v1/status/buildinfo
- /api/v1/labels
- /api/v1/labels
- /api/v1/label/[^/]+/values
- /api/v1/metadata
- /api/v1/query_exemplars
- /api/v1/rules
- /api/v1/series
- /api/v1/series
- /api/v1/status/tsdb
- /loki/api/v1/query_range
- /loki/api/v1/query
- /loki/api/v1/delete
- /loki/api/v1/label
- /loki/api/v1/label
- /loki/api/v1/labels
- /loki/api/v1/labels
- /loki/api/v1/label/[^/]+/values
- /loki/api/v1/label/[^/]+/values
- /loki/api/v1/series
- /loki/api/v1/series
- /tempo/api/traces/[^/]+
- /api/traces/[^/]+
- /api/traces/[^/]+/json
- /tempo/api/echo
- /api/echo
- /tempo/api/search/tags
- /api/search/tags
- /tempo/api/search/tag/[^/]+/values
- /api/search/tag/[^/]+/values
- /api/v2/search/tag/[^/]+/values
- /tempo/api/search
- /api/search

Websocket connection for qryn-reader service:
- /loki/api/v1/tail

# Configuration options

The required options are marked **bold**