This sounds like a really cool feature. I am not aware of any proposals
for this. Feel free to join our sig-storage meetings to discuss this.
On Fri, Aug 24, 2018 at 9:39 AM Sandeep Bangera
wrote:
> Greetings,
>
>
>
You can try the nfs-client provisioner to dynamically create subdirectories
on a single NFS server:
https://github.com/kubernetes-incubator/external-storage/tree/master/nfs-client
On Fri, Aug 24, 2018 at 9:55 AM 'Tim Hockin' via Kubernetes user discussion
and Q wrote:
> If you pointed them at
Hi Matt,
In Kubernetes 1.11, the volume resizing feature is available in beta:
https://kubernetes.io/docs/concepts/storage/persistent-volumes/#expanding-persistent-volumes-claims.
Changing the PVC size will trigger the resizing of the GCE PD disk. Once
that is complete, then you need to restart
Hi Jerry,
With a Daemonset, all instances share the same PVC, so you need to use some
multi writer storage like NFS.
If you want each instance to get a different volume, then you can use
Statefulset's volumeClaimTemplates.
On Thu, Apr 19, 2018, 21:57 Jerry Hwang wrote:
>
ts
> over StatefulSets when using PDs.
>
> On Tue, Mar 20, 2018 at 2:34 PM, 'Michelle Au' via Kubernetes user
> discussion and Q <kubernetes-users@googlegroups.com> wrote:
>
>> Hi Tyler,
>>
>> StatefulSets will also give you a stable network identity, if you need
>
Hi Tyler,
StatefulSets will also give you a stable network identity, if you need
that. Since GCE PDs can only be mounted in write mode by one node, you
should only have one pod referencing it. It is easy to misconfigure
Deployments to end up in scenarios with multiple pods referencing the same
Also, what version of minikube are you running?
On Tue, Jan 2, 2018 at 2:45 PM, Michelle Au wrote:
> I'm not familiar with the minikube environment, but could you double check
> that the minikube-hostpath external provisioner is running, either as a pod
> or a standalone
I'm not familiar with the minikube environment, but could you double check
that the minikube-hostpath external provisioner is running, either as a pod
or a standalone process?
On Tue, Jan 2, 2018 at 12:04 PM, wrote:
> PVC describe -> "waiting for a volume to
If you describe the PVC, what events do you see?
On Sun, Dec 31, 2017 at 8:04 AM, wrote:
> Hi all!
>
> I am trying to set up Cassandra using Kubernetes and came across this
> guide from kubernetes.io:
>
>
Can you paste more details about your PersistentVolume?
On Mon, Dec 18, 2017 at 6:25 PM, wrote:
> Could you please help me to solve this issue, it works fine in minicube
> but fails to mount PVC in linux:
> bash-4.2$ kubectl describe pods/coordinator-1030109887-ptn6l
>
Hi,
Local volumes is an alpha feature currently, so some extra steps are needed
to get it working. I would be happy to help you out on slack.
It is designed for use with stateful sets and provides data gravity by
always scheduling your pod to the correct node.
It also works much better than
What does 'kubectl describe' on the PVC show?
On Thu, Aug 24, 2017 at 6:21 PM, wrote:
> I am trying to create a dynamic storage claim. After deploying my
> StorageClass and PersistentVolumeClaim, the EBS volume is actually created
> and appears in my AWS management
For the RocksDB pod, hostpath should work. Hostpath volumes are not
automatically cleaned up, so it should be able to work for your use case.
Note that you will need to specifically use a node selector in your pods if
you want to access that particular DB instance. If you want a more
scalable
If in your PVC, you:
1. Specify a storage class: Will try to find an existing PV with the same
storageclass. If none exists, it will try to dynamically provision one.
2. Don't specify a storage class: If you have a default storageclass in
your cluster, it will dynamically provision one.
efault" annotation is set?
You will have to change that to false to disable it.
On Mon, May 22, 2017 at 11:36 PM, Tomasz Kapek <kapek...@gmail.com> wrote:
> Thank you Michelle, I check the idea not using classes. It sounds like a
> solution for me ;)
>
> pon., 22 maj 2017, 19
something.
>
> Thanks,
> Arve
>
> On Mon, May 22, 2017 at 7:52 PM 'Michelle Au' via Kubernetes user
> discussion and Q <kubernetes-users@googlegroups.com> wrote:
>
>> Hi Arve,
>>
>> I am not familiar with Rook so I cannot compare. Maybe someone else who
>>
Rook <https://rook.io/> for the same purpose?
>
> Best,
> Arve
>
> On Mon, May 22, 2017 at 7:29 PM 'Michelle Au' via Kubernetes user
> discussion and Q <kubernetes-users@googlegroups.com> wrote:
>
>> Until dynamic provisioning for DigitalOcean is supported,
It looks like you have a number of user-installed storage classes. If you
don't want to use storage classes at all, then you should delete all of
them. Otherwise, in your PVC, specify "" as the storageclass name to bind
to pre-existing PVs that don't have a storage class.
On Sun, May 21, 2017
Yeah, unfortunately you'll have to write a custom operator or scheduler
until local PVs are available. Another option is to consider running as a
DaemonSet, then you will only have one pod per node and won't need a custom
scheduler.
On Mon, May 8, 2017 at 12:33 AM, vincent gromakowski <
We're working on a feature to allow local volumes to be specified as PVs
wth scheduler integration. But it probably won't be in any
production-ready state until at least 1.8.
In the meantime, you might be able to use pod anti-affinity to ensure that
the pods in your statefulset do not end up on
Hi Richard,
Are you sharing the local SSD between many pods, or just one pod per SSD?
If sharing is ok, then in the short term we could look into one of the
following approaches:
1. Ability to create a GKE cluster with kubelet installed on top of a
PD-SSD. Then all EmptyDirs will use this PD.
21 matches
Mail list logo