Reproduce in OCP version:
openshift v3.4.1.18
kubernetes v1.4.0+776c994
etcd 3.1.0-rc.0

geliu I am unable to reproduce this. 

All pods keep running just fine for me after `oc rollout pause dc/<dc-name>. 

Please provide detailed instructions on how to reproduce it if you want to keep it open.

@tnozicka, I reproduce it again on newest 3.4 OCP,

openshift v3.4.1.32
kubernetes v1.4.0+776c994
etcd 3.1.0-rc.0


Steps:

0. Deployment conf file: hello.yaml

apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: hello-openshift
spec:
  replicas: 2
  template:
    metadata:
      labels:
        app: hello-openshift
    spec:
      containers:
      - name: hello-openshift
        image: openshift/hello-openshift
        ports:
        - containerPort: 80
  strategy:
    rollingUpdate:
      maxSurge: 1
      maxUnavailable: 1
    type: RollingUpdate

1. # oc create -f hello.yaml 
deployment "hello-openshift" created
# oc get pods
NAME                               READY     STATUS    RESTARTS   AGE
hello-openshift-2049296760-czvlh   1/1       Running   0          5s
hello-openshift-2049296760-wqeue   1/1       Running   0          5s

2. # oc edit deployment
deployment "hello-openshift" edited

change: "- image: openshift/hello-openshift" to "- image: openshift/nonexist"

3. # oc get pods
NAME                               READY     STATUS             RESTARTS   AGE
hello-openshift-2049296760-wqeue   1/1       Running            0          48s
hello-openshift-2758264543-m7i3l   0/1       ImagePullBackOff   0          19s
hello-openshift-2758264543-x728d   0/1       ImagePullBackOff   0          19s

4. # oc rollout pause deployment/hello-openshift
deployment "hello-openshift" paused
# oc get pods
NAME                               READY     STATUS             RESTARTS   AGE
hello-openshift-2758264543-m7i3l   0/1       ImagePullBackOff   0          32s
hello-openshift-2758264543-x728d   0/1       ImagePullBackOff   0          32s

5. # oc rollout  resume deployment/hello-openshift
deployment "hello-openshift" resumed
[root@host-8-174-99 tmp]# oc get pods
NAME                               READY     STATUS             RESTARTS   AGE
hello-openshift-2758264543-m7i3l   0/1       ImagePullBackOff   0          2m
hello-openshift-2758264543-x728d   0/1       ImagePullBackOff   0          2m

The fix is included in v3.4.1.33

@mkargaki, 3.4.1.33 is not available in the repos: http://download-node-02.eng.bos.redhat.com/rcm-guest/puddles/RHAOS/AtomicOpenShift-signed/3.4/, so I will verify it after it ready, thx

Verified on OCP env:

openshift v3.3.1.35
kubernetes v1.3.0+52492b4
etcd 2.3.0+git

Sorry for typo at comment8, Verified it on OCP 3.4 env:

# openshift version
openshift v3.4.1.33
kubernetes v1.4.0+776c994
etcd 3.1.0-rc.0

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHBA-2017:3049