Ruby-ex deployment times out

2016-09-27 Thread Gerard Braad
Hi All, On a test environment using the `oc cluster up` command, I am unable to deploy the ruby-ex example application. I am deploying on Fedora 24, using origin v1.3. As you can see, the build succeeds, pushed the image and then deployment times out. $ oc logs bc/ruby-ex Cloning "https://githu

Re: Ruby-ex deployment times out

2016-09-27 Thread Clayton Coleman
Were there any logs from the ruby pod as it was being spun up? If you trigger a new deployment: oc deploy ruby-ex --latest When you see a new ruby-ex-* pod get created (the one with the random suffix) try checking the logs with oc logs POD_NAME To see if the app was failing to load. O

Re: Ruby-ex deployment times out

2016-09-27 Thread Gerard Braad
On Tue, Sep 27, 2016 at 11:17 PM, Clayton Coleman wrote: > oc logs POD_NAME $ oc log ruby-ex-2-7ywo7 W0927 15:20:50.019027 17417 cmd.go:269] log is DEPRECATED and will be removed in a future version. Use logs instead. Error from server: container "ruby-ex" in pod "ruby-ex-2-7ywo7" is waitin

Re: Ruby-ex deployment times out

2016-09-27 Thread Clayton Coleman
You can run "oc get events" to see the message about pulling, but it's possible that your machine can't pull from the registry that the image is on. That could be a lot of things - if you're doing a build that pushed an image to the internal registry, docker has to have --insecure-registry set. I

Re: Ruby-ex deployment times out

2016-09-27 Thread Gerard Braad
Hi, although the image was pushed during the build, it says: Failed to pull image "172.30.165.95:5000/myproject/ruby-ex Manually pulling gives: $ docker pull 172.30.165.95:5000/myproject/ruby-ex:latest Trying to pull repository 172.30.165.95:5000/myproject/ruby-ex ... Pulling repository 172.30

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
Hi, On Wed, Sep 28, 2016 at 9:00 AM, Gerard Braad wrote: > Full output I posted here: > https://gist.github.com/gbraad/e82edffb671a5dd154a939491514f7f8 Created an almost similar environment (using CentOS 7 1602 instead of Fedora 24) and the same steps yield a working situation. It is most likely

Re: Ruby-ex deployment times out

2016-09-28 Thread Michail Kargakis
"Cannot overwrite digest sha256:050a263ccf9db73bf688f22f7fdcb690d588dbd1cc8c5a7f50b461ef64dcc88d" Why do we need to do this? On Wed, Sep 28, 2016 at 10:02 AM, Gerard Braad wrote: > Hi, > > On Wed, Sep 28, 2016 at 9:00 AM, Gerard Braad wrote: > > Full output I posted here: > > https://gist.gith

Re: Ruby-ex deployment times out

2016-09-28 Thread Andy Goldstein
Gerard, were you on docker 1.8 and did you then upgrade to 1.9 or 1.10? I recall a bug we had that sounds similar to what you're seeing here. On Wednesday, September 28, 2016, Michail Kargakis wrote: > "Cannot overwrite digest sha256:050a263ccf9db73bf688f22f7fdcb6 > 90d588dbd1cc8c5a7f50b461ef64d

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
On Wed, Sep 28, 2016 at 6:50 PM, Andy Goldstein wrote: > Gerard, were you on docker 1.8 and did you then upgrade to 1.9 or 1.10? I > recall a bug we had that sounds similar to what you're seeing here. I had not done an upgrade at all. a clean cloud image install: $ dnf install -y curl docker $ d

Re: Ruby-ex deployment times out

2016-09-28 Thread Andy Goldstein
Can you run this command: sudo cat /var/lib/docker/image/devicemapper/repositories.json | python -mjson.tool and find the repository entry for 172.30.165.95:5000/myproject/ruby-ex. Then find the entry for http://172.30.165.95:5000/myproject/ruby-ex@sha256:df71f696941a9daa5daaea808cfcaaf72071d7ad2

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
On Wed, Sep 28, 2016 at 10:26 PM, Andy Goldstein wrote: > sudo cat /var/lib/docker/image/devicemapper/repositories.json | python > -mjson.tool https://gist.github.com/gbraad/e82edffb671a5dd154a939491514f7f8#file-to-andy-goldstein-md Currently recreating another environment. If you need access to

Re: Ruby-ex deployment times out

2016-09-28 Thread Michail Kargakis
I just updated from docker 1.9 to docker 1.10 and can reproduce this issue. I managed to deploy by using the RepoTags instead of the RepoDigest. Is this a known issue in the sense that there is a fix to it or we can only pull by tags? On Wed, Sep 28, 2016 at 4:26 PM, Andy Goldstein wrote: > Can

Re: Ruby-ex deployment times out

2016-09-28 Thread Andy Goldstein
The data you pasted in the gist don't match the 172.30.165.95 IP. Are you experiencing the same error message about overwriting the digest with the data in the gist? On Wed, Sep 28, 2016 at 10:50 AM, Gerard Braad wrote: > On Wed, Sep 28, 2016 at 10:26 PM, Andy Goldstein > wrote: > > sudo cat /v

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
On Wed, Sep 28, 2016 at 10:55 PM, Andy Goldstein wrote: > The data you pasted in the gist don't match the 172.30.165.95 IP. Are you > experiencing the same error message about overwriting the digest with the > data in the gist? reproducible... The IP is different because of successive `oc cluste

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
On Wed, Sep 28, 2016 at 10:56 PM, Michail Kargakis wrote: > I just updated from docker 1.9 to docker 1.10 and can reproduce this issue. > I managed to deploy by using the RepoTags instead of the RepoDigest. Is this > a known issue in the sense that there is a fix to it or we can only pull by > tag

Re: Ruby-ex deployment times out

2016-09-28 Thread Michail Kargakis
You can work around this by pulling by tag instead of by hash. Try setting {DOCKER-REGISTRY-IP}/myproject/ruby-ex:latest instead of {DOCKER-REGISTRY-IP}/myproject/ruby-ex@sha256:... in your DC. Not sure how pull by hash can be fixed though. On Wed, Sep 28, 2016 at 5:08 PM, Gerard Braad wrote: >

Re: Ruby-ex deployment times out

2016-09-28 Thread Gerard Braad
On Wed, Sep 28, 2016 at 11:22 PM, Michail Kargakis wrote: > You can work around this by pulling by tag instead of by hash. Try setting > {DOCKER-REGISTRY-IP}/myproject/ruby-ex:latest instead of > {DOCKER-REGISTRY-IP}/myproject/ruby-ex@sha256:... in your DC. > Not sure how pull by hash can be fixed

Re: Ruby-ex deployment times out

2016-09-29 Thread Andy Goldstein
cc Seth, who has done some more investigation and will reply with an update. Andy On Wed, Sep 28, 2016 at 11:30 AM, Gerard Braad wrote: > On Wed, Sep 28, 2016 at 11:22 PM, Michail Kargakis > wrote: > > You can work around this by pulling by tag instead of by hash. Try > setting > > {DOCKER-REG

Re: Ruby-ex deployment times out

2016-09-29 Thread Seth Jennings
Hey Gerard, I also hit this issue this morning: https://paste.fedoraproject.org/438222/51651371/ Seems that a recent change to docker (projectatomic/docker) introduced a regression: $ git describe v1.10.3-95-g8b7fa4a $ git log --oneline a612434.. | grep -i digest 8e24cc3 Merge pull request #194

Re: Ruby-ex deployment times out

2016-09-29 Thread Gerard Braad
On Fri, Sep 30, 2016 at 1:16 AM, Seth Jennings wrote: > You can go to koji and downgrade to an old version (this is the one I used): > http://koji.fedoraproject.org/koji/buildinfo?buildID=796377 Currently on C7, and waiting for a fix instead > I've opened a bug to track this: > https://bugzilla.

Re: Ruby-ex deployment times out

2016-09-30 Thread Gerard Braad
Same is reproducible on Atomic after an rpm-ostree upgrade [fedora@atomic openshift-origin-client-tools-v1.3.0-3ab7af3d097b57f933eccef684a714f2368804e7-linux-64bit]$ docker version Client: Version: 1.10.3 API version: 1.22 Package version: docker-1.10.3-52.git8b7fa4a.fc24.x86_64 G