Chris Watts's questions

I've heard that in later versions of Kubernetes (1.9 onwards if I'm not mistaken; I have 1.10), it's possible to expand a PersistentVolume as long as allowVolumeExpansion: true is set in the StorageClass configuration.

In my case, on GCP, the StorageClass my PVC uses does not have that line, and I cannot add it either.

apiVersion: storage.k8s.io/v1
kind: StorageClass
metadata:
  annotations:
    storageclass.beta.kubernetes.io/is-default-class: "true"
  creationTimestamp: 2018-05-30T17:07:33Z
  labels:
    addonmanager.kubernetes.io/mode: EnsureExists
    kubernetes.io/cluster-service: "true"
  name: standard
  resourceVersion: "8741704"
  selfLink: /apis/storage.k8s.io/v1/storageclasses/standard
  uid: f1bd0421-642b-11e8-bb11-42010a9a00b5
parameters:
  type: pd-standard
provisioner: kubernetes.io/gce-pd
reclaimPolicy: Delete
volumeBindingMode: Immediate

So, I want to increase a PV, and its corresponding PVC from 8Gi to 100Gi. What's the best way of doing this? Is there a way to do it while preserving data, or must the current PV be deleted before making a new one?

Here's the PV YAML:

apiVersion: v1
kind: PersistentVolume
metadata:
  annotations:
    kubernetes.io/createdby: gce-pd-dynamic-provisioner
    pv.kubernetes.io/bound-by-controller: "yes"
    pv.kubernetes.io/provisioned-by: kubernetes.io/gce-pd
  creationTimestamp: 2018-05-31T10:30:39Z
  finalizers:
  - kubernetes.io/pv-protection
  labels:
    failure-domain.beta.kubernetes.io/region: europe-west2
    failure-domain.beta.kubernetes.io/zone: europe-west2-c
  name: pvc-a70ca000-64bd-11e8-bb11-42010a9a00b5
  resourceVersion: "8728415"
  selfLink: /api/v1/persistentvolumes/pvc-a70ca000-64bd-11e8-bb11-42010a9a00b5
  uid: a9e8c071-64bd-11e8-bb11-42010a9a00b5
spec:
  accessModes:
  - ReadWriteOnce
  capacity:
    storage: 8Gi
  claimRef:
    apiVersion: v1
    kind: PersistentVolumeClaim
    name: production-postgres
    namespace: neserver-6540663
    resourceVersion: "85487"
    uid: a70ca000-64bd-11e8-bb11-42010a9a00b5
  gcePersistentDisk:
    fsType: ext4
    pdName: gke-cluster-1-320626e3-pvc-a70ca000-64bd-11e8-bb11-42010a9a00b5
  persistentVolumeReclaimPolicy: Delete
  storageClassName: standard
status:
  phase: Bound

When using MIT Kerberos Ticket Manager with PuTTY 0.65 and WinSCP 5.9.3, I am sometimes unable to get a connnection to the server I am logging into. PuTTY will respond with either No supported authentication methods available (server sent: ) or No supported authentication methods available (server sent: publickey,gssapi-keyex,gssapi-with-mic)

It seems to happen most often if I try to connect to the server before remembering to obtain a ticket first. After getting a ticket, I will keep getting the same error every time I try. It sometimes resolves itself after about an hour.

The connection to both the authentication server and the server I'm trying to connect to operates over a VPN.

I don't get any issues when using Cygwin + kinit.

Here's my log:

. 2017-01-28 14:04:38.979 Server offered these authentication methods: publickey,gssapi-keyex,gssapi-with-mic
. 2017-01-28 14:04:38.980 Using SSPI from SECUR32.DLL
. 2017-01-28 14:04:38.980 Attempting GSSAPI authentication
. 2017-01-28 14:04:39.008 GSSAPI authentication initialisation failed
. 2017-01-28 14:04:39.008 No authority could be contacted for authentication.The domain name of the authenticating party could be wrong, the domain could be unreachable, or there might have been a trust relationship failure.
. 2017-01-28 14:04:39.008 Disconnected: No supported authentication methods available (server sent: )
* 2017-01-28 14:04:39.060 (EFatal) Disconnected: No supported authentication methods available (server sent: )

I'm trying to daemonize a process under a different user.

In my init.d, I have a service which looks a little bit like this:

...
start() {
    echo "Starting mydaemon..."
    daemon --user someuser --name mydaemon mycommand
}
...

(entire script here: http://pastebin.com/bvpnsHgn)

However, all I get is daemon: command not found.

I checked that daemon() is declared in /etc/init.d/functions, but still, I cannot seem to use it.

How can I resolve this?

In my current CentOS ifconfig, I have eth0 and the range of virtual interfaces eth0:0 - eth0:3 each taking up a unique address in the subnet 69.xx.yy.194/29 (gateway = .193, broadcast = .199)

Now I only get one shot at this because I access the server via remote terminal, so I thought I'd ask here to double check what I'm about to do.

I'm guessing if I perform ifdown eth0, it will take everything, including the interfaces down. What I would like is to be able to take down any one of my 5 services running on the machine by plugging them each into their own virtual interface (note, I only have 4 interfaces right now)

Therefore, is it possible to not have an IP address assigned to eth0, and rather put it on eth0:0 so I end up with 5 interfaces each with a separate address?

(Will this have unexpected side-effects for my PuTTY connection?)

I've made a remote git repository with

mkdir /var/www/html/myproject.git
chmod 0770 /var/www/html/myproject.git
ln -s /var/www/html/myproject.git myproject.git
cd myproject.git
git --bare init

and successfully connected to it and cloned it on my own machine with

git clone [email protected]:myproject.git

then uploaded some files to it

git push

which when tried again shows that the repository is already up to date.

However, I'm looking around on the remote server, and can't actually find where git has stored those files! It's not in /var/www/html/ nor /home/git and the only thing I've managed to find is the HEAD file which contains the id of the initial commit I made in /var/www/html/myproject.git/refs/heads/master. (So I know the push worked)

Where could the files possibly be?