[bitnami/mongodb] Can't communicate with replicaset using external access

[cmanzur]

Name and Version

bitnami/mongodb 13.3.1

What steps will reproduce the bug?

I have deployed a mongodb replicaset (v6.0.2) using this chart: bitnami/mongodb 13.3.1 I have enabled external access with SVC type LoadBalancer. I have my DNS records pointing to each external SVC IP.

• mongodb-0.dev -> 10.246.50.1
• mongodb-1.dev -> 10.246.50.2
• mongodb-2.dev -> 10.246.50.3

$ kubectl get pod,svc -owide
NAME                    READY   STATUS               
pod/mongodb-0           2/2     Running   
pod/mongodb-1           2/2     Running  
pod/mongodb-2           2/2     Running   
pod/mongodb-arbiter-0   2/2     Running   

NAME                               TYPE           CLUSTER-IP      EXTERNAL-IP    PORT(S)           AGE     SELECTOR
service/mongodb-0-external         LoadBalancer   10.43.182.101   10.246.50.1   27017:30727/TCP   6m6s    app.kubernetes.io/component=mongodb,app.kubernetes.io/instance=mongodb,app.kubernetes.io/name=mongodb,statefulset.kubernetes.io/pod-name=mongodb-0
service/mongodb-1-external         LoadBalancer   10.43.130.10    10.246.50.2   27017:30137/TCP   6m6s    app.kubernetes.io/component=mongodb,app.kubernetes.io/instance=mongodb,app.kubernetes.io/name=mongodb,statefulset.kubernetes.io/pod-name=mongodb-1
service/mongodb-2-external         LoadBalancer   10.43.86.149    10.246.50.3   27017:32246/TCP   6m6s    app.kubernetes.io/component=mongodb,app.kubernetes.io/instance=mongodb,app.kubernetes.io/name=mongodb,statefulset.kubernetes.io/pod-name=mongodb-2
service/mongodb-arbiter-headless   ClusterIP      None            <none>         27017/TCP         8m51s   app.kubernetes.io/component=arbiter,app.kubernetes.io/instance=mongodb,app.kubernetes.io/name=mongodb
service/mongodb-headless           ClusterIP      None            <none>         27017/TCP         8m51s   app.kubernetes.io/component=mongodb,app.kubernetes.io/instance=mongodb,app.kubernetes.io/name=mongodb

I run a docker from OUTSIDE to test the client:

docker run --rm -it -v docker.io/bitnami/mongodb:6.0.2-debian-11-r1 bash

Option 1: Direct connection from OUTSIDE

I CAN connect using:

mongosh  mongodb://mongodb-0.dev:27017 --authenticationDatabase admin -u root -p root1234
Connecting to:      mongodb://<credentials>@mongodb-0.dev:27017/?directConnection=true&authSource=admin

rs0 [direct: primary] test> 

Option 2: Replica Set Connection from OUTSIDE

I CAN'T connect using:

mongosh  mongodb://mongodb-0.dev:27017,mongodb-1.dev:27017,mongodb-2.dev:27017?replicaSet=rs0 --authenticationDatabase admin -u root -p root1234

Connecting to:      mongodb://<credentials>@mongodb-0.dev:27017,mongodb-1.dev:27017,mongodb-2.dev:27017/?replicaSet=rs0&authSource=admin&appName=mongosh+1.6.0

MongoNetworkError: getaddrinfo ENOTFOUND mongodb-0.mongodb-headless.mongodb.svc.cluster.local

Of course, If I add the DNS records it works (pay attention at the prompt)::

• mongodb-0.mongodb-headless.mongodb.svc.cluster.local -> 10.246.50.1
• mongodb-1.mongodb-headless.mongodb.svc.cluster.local -> 10.246.50.2
• mongodb-2.mongodb-headless.mongodb.svc.cluster.local -> 10.246.50.3

Connecting to:      mongodb://<credentials>@mongodb-0.dev:27017,mongodb-1.dev:27017,mongodb-2.dev:27017/?replicaSet=rs0&authSource=admin

rs0 [primary] test> 

But I don't want to do that workaround at DNS level, of course it's wrong! I also don't want to harcode any IP on /etc/hosts

Option 3: From INSIDE the k8s cluster

This is just extra information in case it helps to find the issue. I CAN connect using:

kubectl run mongo-client-6 --rm -ti --image=docker.io/bitnami/mongodb:6.0.2-debian-11-r1 -- bash
mongosh --host mongodb-headless --authenticationDatabase admin -u root -p root1234

Summary: I want to connect using the Option 2 (replica set from OUTSIDE)

Are you using any custom parameters or values?

architecture: replicaset
replicaCount: 3

externalAccess:
  enabled: true
  autoDiscovery:
    enabled: true

auth:
  enabled: true
  rootUser: root
  rootPassword: "root1234"
  replicaSetKey: "adfaerfeqawfdasefa"

What do you see instead?

MongoNetworkError: getaddrinfo ENOTFOUND mongodb-0.mongodb-headless.mongodb.svc.cluster.local

[fmulero]

Hi Carim

Have you tried upgrading your chart setting externalAccess.service.loadBalancerIPs with the external IPs?

[cmanzur]

Hi Carim

Have you tried upgrading your chart setting externalAccess.service.loadBalancerIPs with the external IPs?

Yes, I tried everything. It didn't work.

[fmulero]

I was trying to reproduce your issue in version 13.3.1 but I faced some issues (#13201) already fixed in latest version (13.5.0 at the time of this comment). Could you try latest version enabling rbac for autodiscovery?

architecture: replicaset
replicaCount: 3

rbac:
  create: true
externalAccess:
  enabled: true
  autoDiscovery:
    enabled: true

auth:
  enabled: true
  rootUser: root
  rootPassword: "root1234"
  replicaSetKey: "adfaerfeqawfdasefa"

[github-actions[bot]]

This Issue has been automatically marked as "stale" because it has not had recent activity (for 15 days). It will be closed if no further activity occurs. Thanks for the feedback.

[github-actions[bot]]

Due to the lack of activity in the last 5 days since it was marked as "stale", we proceed to close this Issue. Do not hesitate to reopen it later if necessary.

[riaanp78]

@cmanzur did you ever get this to work externally? I have the same problem as you. works 100% inside Kubernetes, but external access does not. I get these errors:

Name or service not known,mongodb5-2.mongodb5-headless.mongodb-dev.svc.cluster.local:27017: [Errno -2]

So from the looks of it will never word externally unless you hack the /etc/hosts

[fabrizziocht]

@cmanzur did you ever get this to work externally? I have the same problem as you. works 100% inside Kubernetes, but external access does not. I get these errors:

Name or service not known,mongodb5-2.mongodb5-headless.mongodb-dev.svc.cluster.local:27017: [Errno -2]

So from the looks of it will never word externally unless you hack the /etc/hosts

Hello, I was trying to connect enabling external access and rbac settiings, so when I include the parameter directConnection in the connection string I can connect locally: mongodb://root:root@localhost:27017/?directConnection=true Regards, Fabri

[igolka97]

The issue still persists. I found out that replicaset initialised with names like 192.168.0.7:30001 (where 192.168.0.7 is my internal host ip and 30001 is nodePort. It possible to connect externally with directConnection=true setting FQDN or ip only once. In case of that, when you try to connect to cluster without directConnection your external client trying to connect to all replicas in set at once right after connection init and fails, as of 192.168.0.7 ip available only in cluster network. I guess that MONGODB_ADVERTISED_HOSTNAME and following lines is a culprit: https://github.com/bitnami/charts/blob/973a2792e0bc5967e3180c6d44eebf223b9f1d83/bitnami/mongodb/templates/replicaset/statefulset.yaml#L202-L205

values:

architecture: replicaset
replicaCount: 2
externalAccess:
  enabled: true
  service:
    type: NodePort
    port: 27017
    nodePorts:
      - 30001
      - 30002
    publicNames:
      - sub.domain.com
      - sub.domain.com
    domain: "sub.domain.com"
    externalTrafficPolicy: Cluster

members: [
    {
      _id: 0,
      name: '192.168.0.7:30001',
      health: 1,
      state: 1,
      stateStr: 'PRIMARY',
      uptime: 586,
      optime: [Object],
      optimeDate: 2024-09-28T22:21:37.000Z,
      lastAppliedWallTime: 2024-09-28T22:21:37.619Z,
      lastDurableWallTime: 2024-09-28T22:21:37.619Z,
      syncSourceHost: '',
      syncSourceId: -1,
      infoMessage: '',
      electionTime: Timestamp({ t: 1727561547, i: 1 }),
      electionDate: 2024-09-28T22:12:27.000Z,
      configVersion: 4,
      configTerm: 8,
      self: true,
      lastHeartbeatMessage: ''
    },
    {
      _id: 1,
      name: '192.168.0.6:30002',
      health: 1,
      state: 2,
      stateStr: 'SECONDARY',
      uptime: 562,
      optime: [Object],
      optimeDurable: [Object],
      optimeDate: 2024-09-28T22:21:37.000Z,
      optimeDurableDate: 2024-09-28T22:21:37.000Z,
      lastAppliedWallTime: 2024-09-28T22:21:37.619Z,
      lastDurableWallTime: 2024-09-28T22:21:37.619Z,
      lastHeartbeat: 2024-09-28T22:21:39.875Z,
      lastHeartbeatRecv: 2024-09-28T22:21:40.286Z,
      pingMs: Long('0'),
      lastHeartbeatMessage: '',
      syncSourceHost: '192.168.0.7:30001',
      syncSourceId: 0,
      infoMessage: '',
      configVersion: 4,
      configTerm: 8
    }
  ],```

[github-actions[bot]]

This Issue has been automatically marked as "stale" because it has not had recent activity (for 15 days). It will be closed if no further activity occurs. Thanks for the feedback.

[github-actions[bot]]

Due to the lack of activity in the last 5 days since it was marked as "stale", we proceed to close this Issue. Do not hesitate to reopen it later if necessary.