I see this happening also with podman-1.6.4-11.module+el8.3.0+6219+ce16cb0a.x86_64
And confirming that it was not happening with podman from 8.1.0.
+++ This bug was initially created as a clone of Bug #1754922 +++
Description of problem:
The man podman-exec(1) says
--interactive, -i=true|false
When set to true, keep stdin open even if not attached. The default is false.
However, I am not able to pipe to podman to get the data to the process running in the container via podman exec.
Version-Release number of selected component (if applicable):
podman-1.5.1-3.git0005792.fc31.x86_64
podman-1.5.2-0.27.dev.git112a3cc.fc32.x86_64
How reproducible:
Deterministic.
Steps to Reproduce:
1. podman run --rm --name test-podman-exec -d registry.fedoraproject.org/fedora:30 sleep 60
2. echo asdf | podman exec -i test-podman-exec cat
Actual results:
On Fedora 31: empty output.
On Fedora rawhide:
Error: read unixpacket @->/var/run/libpod/socket/59ad86c6f892d4b02a76e240615b1085995bba891072fb1895e28f193b3883b5/attach: read: connection reset by peer
Expected results:
asdf
Additional info:
This is a regression against podman-1.4.4-4.fc30.x86_64.
We have freeipa-container tests where we run
echo $PASSWORD | sudo podman exec -i the-container kinit admin
It started to fail with podman (1.5.1-1~ubuntu16.04~ppa1). I was able to reproduce the problem on Fedora, and on rawhide it seems even more broken than in Fedora 31.
--- Additional comment from Jan Pazdziora on 2019-09-24 11:28:27 UTC ---
I've also tested podman exec -i=true and --interactive, in case they were needed to make the -i option actually do something (unlike docker and unlike podman 1.4).
The podman exec -i=true behaves exactly like -i on both Fedora 31 (no output) and on Fedora rawhide (that connection reset by peer error).
However, podman exec --interactive causes even Fedora 31 to produce that connection reset by peer error.
--- Additional comment from Matthew Heon on 2019-10-28 14:08:51 UTC ---
Upstream tracker: https://github.com/containers/libpod/issues/3302
This one is proving extremely difficult to track down, but seems to be in the handling of `-i` with exec after we migrated to conmon.
--- Additional comment from Ben Cotton on 2019-10-31 18:43:42 UTC ---
This message is a reminder that Fedora 29 is nearing its end of life.
Fedora will stop maintaining and issuing updates for Fedora 29 on 2019-11-26.
It is Fedora's policy to close all bug reports from releases that are no longer
maintained. At that time this bug will be closed as EOL if it remains open with a
Fedora 'version' of '29'.
Package Maintainer: If you wish for this bug to remain open because you
plan to fix it in a currently maintained version, simply change the 'version'
to a later Fedora version.
Thank you for reporting this issue and we are sorry that we were not
able to fix it before Fedora 29 is end of life. If you would still like
to see this bug fixed and are able to reproduce it against a later version
of Fedora, you are encouraged change the 'version' to a later Fedora
version prior this bug is closed as described in the policy above.
Although we aim to fix as many bugs as possible during every release's
lifetime, sometimes those efforts are overtaken by events. Often a
more recent Fedora release includes newer upstream software that fixes
bugs or makes them obsolete.
--- Additional comment from Jan Pazdziora on 2019-11-01 09:34:58 UTC ---
Updating version as the issue is still present with podman-1.6.2-2.fc31.x86_64.
--- Additional comment from Daniel Walsh on 2019-11-01 13:31:26 UTC ---
Peter do the fixes we just merged into podman fix this issue?
--- Additional comment from Peter Hunt on 2019-11-01 13:41:49 UTC ---
No, the piped input still doesn't seem to make it to the container, but we're getting closer. Work still needs to be done
--- Additional comment from Peter Hunt on 2020-01-15 19:50:17 UTC ---
Finally found a fix in https://github.com/containers/libpod/pull/4818
This will land in either 1.7.1 or 1.8 depending on which version is cut next