Skip to content

How to use a Spaces Object Storage bucket as RWX storage for a Kubernetes Pod with DOKS (DigitalOcean Kubernetes)

Notifications You must be signed in to change notification settings

DO-Solutions/doks-s3

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 

Repository files navigation


Logo

How to use a Spaces Object Storage bucket as ReadWriteMany (RWX) storage for a Kubernetes Pod with DOKS (DigitalOcean Kubernetes)

Introduction

  • In this guide, we will deploy a DigitalOcean Managed Kubernetes cluster.
  • We'll use k8s-csi-s3, which is a GeeseFS-based CSI for mounting S3 buckets as PersistentVolumes and give some working examples of consuming ReadWriteMany RMX storage.

About DigitalOcean DOKS

DigitalOcean Kubernetes (DOKS) is a managed Kubernetes service that lets you deploy Kubernetes clusters without the complexities of handling the control plane and containerized infrastructure. Clusters are compatible with standard Kubernetes toolchains and integrate natively with DigitalOcean Load Balancers and volumes.

About DigitalOcean Spaces Object Storage

Spaces Object Storage is an S3-compatible object storage service that lets you store and serve large amounts of data. Each Space is a bucket for you to store and serve files. The built-in Spaces CDN minimizes page load times and improves performance.

About GeeseFS

Finally, an exemplary FUSE FS implementation over S3

  • GeeseFS allows you to mount an S3 bucket as a file system.
  • FUSE file systems based on S3 typically have performance problems, especially with small files and metadata operations.
  • GeeseFS attempts to solve these problems by using aggressive parallelism and asynchrony.
  • GeeseFS is stable enough to pass most applicable xfstests, including dirstress/fsstress stress-tests (generic/007, generic/011, generic/013).
  • Benchmarks compared to rclone+cache, goofys and s3fs: https://github.com/yandex-cloud/geesefs/tree/master/bench

Architecture diagram

Architecture diagram

Prerequisites

  1. A DigitalOcean account (Log in)
  2. A DigitalOcean Spaces Object Storage subscription
  3. doctl CLI(tutorial)

Deploy a DigitalOcean DOKS Cluster

How to create a Kubernetes cluster using the DigitalOcean CLI

To create a Kubernetes cluster via the command line, follow these steps:

  1. Install doctl, the DigitalOcean command-line tool.

  2. Create a personal access token, and save it with doctl.

  3. Use the token to grant doctl access to your DigitalOcean account.

    doctl auth init
  4. Finally, create a Kubernetes cluster with doctl kubernetes cluster create.

    doctl kubernetes cluster create doks-shark-1 --auto-upgrade=true --ha=true --node-pool="name=pool-apps;size=s-4vcpu-8gb-amd;count=3" --region=ams3 --surge-upgrade=true
  • doks-shark-1 is our cluster name.
  • We are creating one node pool for our regular workloads.
  • The region is Amsterdam; surge-upgrades are enabled, HA is enabled, and auto-upgrade is enabled.
  • You'll want to read the usage docs for more details

Setup k8s-csi-s3

Clone the k8s-csi-s3 repo locally

git clone https://github.com/yandex-cloud/k8s-csi-s3

1. Create a secret with your S3 credentials

Replace the endpoint URL with the same region as your DOKS cluster.
Spaces availability per region is detailed here.

apiVersion: v1
kind: Secret
metadata:
  name: csi-s3-secret
  namespace: kube-system
stringData:
  accessKeyID: <YOUR_ACCESS_KEY_ID>
  secretAccessKey: <YOUR_SECRET_ACCESS_KEY>
  endpoint: https://ams3.digitaloceanspaces.com

2. Deploy the driver

cd deploy/kubernetes
kubectl create -f provisioner.yaml
kubectl create -f driver.yaml
kubectl create -f csi-s3.yaml

3. Create the storage class

kubectl create -f examples/storageclass.yaml

4. Test the S3 driver

  1. Create a pvc using the new storage class:

    kubectl create -f examples/pvc.yaml
  2. Check if the PVC has been bound:

    $ kubectl get pvc csi-s3-pvc
    NAME         STATUS    VOLUME                                     CAPACITY   ACCESS MODES   STORAGECLASS   AGE
    csi-s3-pvc   Bound     pvc-c5d4634f-8507-11e8-9f33-0e243832354b   5Gi        RWO            csi-s3         9s
  3. Create a test pod that mounts your volume:

    kubectl create -f examples/pod.yaml

    If the pod can start, everything should be working.

  4. Test the mount

    $ kubectl exec -ti csi-s3-test-nginx bash
    $ mount | grep fuse
    pvc-035763df-0488-4941-9a34-f637292eb95c: on /usr/share/nginx/html/s3 type fuse.geesefs (rw,nosuid,nodev,relatime,user_id=65534,group_id=0,default_permissions,allow_other)
    $ touch /usr/share/nginx/html/s3/hello_world

If something does not work as expected, check the troubleshooting section below.

Additional configuration

Bucket

By default, csi-s3 will create a new bucket per volume. The bucket name will match that of the volume ID. If you want your volumes to live in a pre-created bucket, you can specify the bucket in the storage class parameters:

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: csi-s3-existing-bucket
provisioner: ru.yandex.s3.csi
parameters:
  mounter: geesefs
  options: "--memory-limit 1000 --dir-mode 0777 --file-mode 0666"
  bucket: some-existing-bucket-name

If the bucket is specified, it will still be created if it does not exist on the backend. Every volume will get its prefix within the bucket, which matches the volume ID. When deleting a volume, just the prefix will be deleted.

Static Provisioning

You can use static provisioning if you want to mount a pre-existing bucket or prefix within a pre-existing bucket and don’t want csi-s3 to delete it when PV is deleted.

To do that, you should omit storageClassName in the PersistentVolumeClaim and manually create a PersistentVolume with a matching claimRef, like in the following example: deploy/kubernetes/examples/pvc-manual.yaml.

You can check the POSIX compatibility matrix here: https://github.com/yandex-cloud/geesefs#posix-compatibility-matrix.

GeeseFS

  • Almost full POSIX compatibility
  • Good performance for both small and big files
  • Does not store file permissions and custom modification times
  • By default runs outside of the csi-s3 container using systemd, to not crash mountpoints with "Transport endpoint is not connected" when csi-s3 is upgraded or restarted. Add --no-systemd to parameters.options of the StorageClass. to disable this behavior.

Troubleshooting

Issues while creating PVC

Check the logs of the provisioner:

kubectl logs -l app=csi-provisioner-s3 -c csi-s3 -n kube-system

Issues creating containers

  1. Check the logs of the s3-driver:
kubectl logs -l app=csi-s3 -c csi-s3 -n kube-system

Benchmarks

Spaces Object Storage limits are detailed here

Tests using dd and fio

  • Use gen_small.py to create 6400 files, sized 0.5-300KB, 30KB on average, sharded over 1024 dirs with 2 level deep nesting
    • Copy this directory
    • Delete this directory
  • Write 1GB and 5GB files to Spaces Object Storage
  • Read 1GB and 5GB files from Spaces Object Storage
Test Command Time Detail
Create 6400 files python3 gen_small.py /mnt/s3/test1 11.3 s
Copy the directory cp -r test1 test2 7.8 s
Delete the directory rm -r test1 1.2 s
Write 1GB dd if=/dev/zero of=largefile bs=1MB count=1000 oflag=direct 6.8215 s 147 MB/s
Read 1GB dd if=largefile of=/dev/null bs=1MB iflag=direct 2.02862 s 493 MB/s
Write 5GB dd if=/dev/zero of=largefile5 bs=1MB count=5000 oflag=direct 56.2905 s 88.8 MB/s
Read 5GB dd if=largefile5 of=/dev/null bs=1MB iflag=direct 7.60369 s 658 MB/s

Dbench

Benchmarks ran using dbench

Native volume benchmarks

Below are the results of an s-2vcpu-4gb-amd worker node with a 1TB Volume attached using the do-block-storage storageClass

==================
= Dbench Summary =
==================
Random Read/Write IOPS: 9986/9987. BW: 384MiB/s / 387MiB/s
Average Latency (usec) Read/Write: 750.36/399.11
Sequential Read/Write: 384MiB/s / 395MiB/s
Mixed Random Read/Write IOPS: 7515/2471

S3 benchmarks

==================
= Dbench Summary =
==================
Random Read/Write IOPS: 126/1431. BW: 15.4MiB/s / 47.2MiB/s
Average Latency (ms) Read/Write: 36.1/3.1
Sequential Read/Write: 64.9MiB/s / 61.2MiB/s
Mixed Random Read/Write IOPS: 46/14

Conclusion

In this guide, we deployed a DigitalOcean Managed Kubernetes cluster. We setup k8s-csi-s3, which is a GeeseFS-based CSI for mounting S3 buckets as PersistentVolumes by deploy the driver, creating a new storage class and deploying a test pod with RWX storage.

References

k8s-csi-s3
geesefs

Contact

Jack Pearce, Solutions Engineer - [email protected]

(back to top)

About

How to use a Spaces Object Storage bucket as RWX storage for a Kubernetes Pod with DOKS (DigitalOcean Kubernetes)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published