Skip to content

Latest commit

 

History

History
150 lines (124 loc) · 7.17 KB

README.md

File metadata and controls

150 lines (124 loc) · 7.17 KB
Raw

Container Storage Interface (CSI) for MooseFs

Container storage interface is an industry standard that will enable storage vendors to develop a plugin once and have it work across a number of container orchestration systems.

MooseFs is an open-source distributed file system which aims to be fault-tolerant, highly available, highly performing, scalable general-purpose network distributed file system for data centers.

Introduction

Similar to other storage providers MooseFs can act as a layer on top of hybrid storage. The storage can be distributed across multiple private/public clouds.

alt MooseFSCSI

How it works (Kubernetes)

MooseFs abstracts heterogenous storage providers and acts as a single interface. Here, you can see a kubernetes cluster with moosefs-csi in GKE having storage from AWS elastic block store and Google's persistant disk. alt k8sMooseFs

Version Compatibility

Kubernetes csi-cluster-driver-registrar csi-node-driver-registrar csi-attacher csi-provisioner MooseFS CSI AWS MooseFS CSI GKE MooseFS CSI Azure MooseFS CSI EP (On_prem)
v1.13.1 v1.0.1 v1.0.1 v1.0.1 v1.0.1 v0.0.4 - - v0.0.4

Deployment (Kubernetes)

Prerequisites:

  • Already have a working Kubernetes cluster (includes kubectl)
  • AWS/GCP/Azure credentials available (not needed for on premise clusters[EP])

Storage deployment topology (Optional)

  • Inorder to work with MooseFS, one should choose the MooseFS topology. MooseFS consists of Master and Chunk servers, get to know more about MooseFS architecture here.
  • The topology to be passed to this plugin should be of format: master:<Provider>,chunk:<Provider>. Some valid examples:
    • master:EP,chunk:EP, EP here stands for endpoint
    • master:AWS,chunk:AWS,
    • master:AWS,chunk:GCP,
    • chunk:AZURE,master:GCP
  • Limitations: As of now, there is support only for: AWS,AWS and EP,EP i.e. the first two example above. The rest of the combinations are under development.

AWS EBS storage for your Kubernetes cluster

  1. Git clone this repository and add your credentials to deploy/kubernetes/moosefs-csi.yaml

  2. Apply the container storage interface for moosefs for your cluster

$ kubectl apply -f deploy/kubernetes/moosefs-csi.yaml
  1. Ensure all the containers are ready and running
$ kubectl get po -n kube-system
  1. Testing: Create a persistant volume claim for 5GiB with name moosefs-csi-pvc with storage class moosefs-block-storage
$ kubectl apply -f deploy/kubernetes/sample-moosefs-pvc.yaml
  1. Verify if the persistant volume claim exists and wait until its the STATUS is Bound
$ kubectl get pvc
  1. After its in Bound state, create a sample workload mounting that volume
$ kubectl apply -f deploy/kubernetes/sample-busybox-pod.yaml
  1. Verify the storage mount of the busybox pod
$ kubectl exec my-csi-app -- df -h
  1. Clean up
$ kubectl delete -f deploy/kubernetes/sample-busybox-pod.yaml
$ kubectl delete -f deploy/kubernetes/sample-moosefs-pvc.yaml
$ kubectl delete -f deploy/kubernetes/moosefs-csi.yaml

Utilize exiting on-premise MooseFS storage for your Kubernetes cluster (bare metal)

  1. Git clone this repository and add your MooseFS master IP to deploy/kubernetes/moosefs-csi-ep.yaml (2 places)

  2. Apply the container storage interface for moosefs for your cluster

$ kubectl apply -f deploy/kubernetes/moosefs-csi-ep.yaml
  1. Ensure all the containers are ready and running
$ kubectl get po -n kube-system
  1. Testing: Create a persistant volume claim for 5GiB with name moosefs-csi-pvc with storage class moosefs-block-storage. The value, 5Gib does not have any significance as for MooseFS the whole filesystem is mounted into the container.
$ kubectl apply -f deploy/kubernetes/sample-moosefs-pvc.yaml
  1. Verify if the persistant volume claim exists and wait until its the STATUS is Bound
$ kubectl get pvc
  1. After its in Bound state, create a sample workload mounting that volume
$ kubectl apply -f deploy/kubernetes/sample-busybox-pod.yaml
  1. Verify the storage mount of the busybox pod
$ kubectl exec my-csi-app -- df -h
  1. Clean up
$ kubectl delete -f deploy/kubernetes/sample-busybox-pod.yaml
$ kubectl delete -f deploy/kubernetes/sample-moosefs-pvc.yaml
$ kubectl delete -f deploy/kubernetes/moosefs-csi.yaml

Debuging

Section Issues Commands
Persistance volume claim pvc is in Pending state kubectl get events
kubectl logs csi-attacher-moosefs-plugin-0 -c moosefs-csi-plugin -n kube-system
kubectl logs csi-provisioner-moosefs-plugin-0 -c moosefs-csi-plugin -n kube-system
Driver name not found MountVolume.MountDevice failed for volume "pvc-XYZ" : driver name com.tuxera.moosefs.csi not found in the list of registered CSI drivers Ensure kube-apiserver and kubelet has the correct feature-gates enabled (for k8s < 1.14)

Developing and contributing

  1. Add and commit your changes after forking this repository
  2. Run unit tests
make test
  1. After its successfull, create a pull request ❤️ ❤️

Miscelleneous

Description Command
AWS session token creation aws sts get-session-token --duration-seconds 129600
Docker command for launching moosefs docker run --cap-add SYS_ADMIN --security-opt apparmor:unconfined -v /dev/fuse:/dev/fuse --privileged -it mfs /bin/bash

TODO

  • Possibility to define the count of physical disks on servers
  • Possibility to define the count of chunk servers
  • Automatically generate the best mooseFs configration based on user's need/metrics (how many chunks, topology of servers etc.)
  • Possibility to define replication goal (Erosure codes)

Known Bugs/limitations

  • AWS
    • Chunk server creation is not idempotent, sometimes creates more chunk servers than requested causing more disk space available than requested.
    • AWS Security Group is not deleted while decommissioning, needs to be done manually
    • Fargate cluster is not deleted while decommissioning

MountVolume.MountDevice failed for volume "pvc-c3738d2cc47711e8" : driver name com.digitalocean.csi.dobs not found in the list of registered CSI drivers

License

Apache v2 license

Code of conduct

Participation in this project is governed by Kubernetes/CNCF code of conduct