DockerClientFactory is wrong at guessing available disk space #223

Activity

1. rnorth commented on Sep 14, 2016

   

   rnorth

   on Sep 14, 2016

   Member

   More actions

   Matthieu,

   Thanks - that's very odd. I've seen the results of df vary in interesting
   ways (seemingly based on host OS and/or volume driver) but never 'empty'
   output like that.

   I'll look to fix, but will also have a bit of a rethink bout the utility of
   performing disk space checks - it was useful in the early days of the
   library when image removal was troublesome, but perhaps now it's not
   serving a useful purpose...

   Richard
   On Tue, 13 Sep 2016 at 16:22, Matthieu Baechler notifications@github.com
   wrote:

   Here is what I get when starting a container :

   [main] INFO org.testcontainers.DockerClientFactory - Connected to docker:
   Server Version: 1.12.1-rc2
   API Version: 1.24
   Operating System: Debian GNU/Linux stretch/sid
   Total Memory: 7783 MB
   [main] INFO org.testcontainers.DockerClientFactory - Disk utilization in Docker environment is 0% (0 MB available)
   [main] ERROR org.testcontainers.DockerClientFactory - Docker environment has less than 2GB free - execution is unlikely to succeed so will be aborted.

   The problem is here IMO : CreateContainerResponse createContainerResponse
   = client.createContainerCmd("alpine:3.2").withCmd("df", "-P").exec();

   The result of such a command is :

   $ docker run --rm -ti alpine:3.2 df -P
   Filesystem 1024-blocks Used Available Capacity Mounted on
   $

   It's certainly what is expected.

   —
   You are receiving this because you are subscribed to this thread.
   Reply to this email directly, view it on GitHub
   #223, or mute
   the thread
   https://github.com/notifications/unsubscribe-auth/AAIET7r1GS2D6QTenxmMUPqhtnoo1TQOks5qpr-dgaJpZM4J7zRr
   .

   Richard
2. mbaechler commented on Sep 15, 2016

   

   mbaechler

   on Sep 15, 2016

   ContributorAuthor

   More actions

   I tried running the same command with a debian-based image and the output is non-empty.
   We could at least prevent failure if we are not able to parse df output, It could be enough for now ?
3. mrVanDalo commented on Sep 16, 2016

   

   mrVanDalo

   on Sep 16, 2016

   More actions

   Have the same problem here :(
4. mbaechler commented on Sep 20, 2016

   

   mbaechler

   on Sep 20, 2016

   ContributorAuthor

   More actions

   #225
5. rnorth commented on Nov 20, 2016

   

   rnorth

   on Nov 20, 2016

   Member

   More actions

   I'll close this as the fix was merged previously.
6. rnorth

   closed this as completedon Nov 20, 2016