glusterfs
glusterfs copied to clipboard
gluster
tests: replace localhost and 127.0.01 with valid ip
Problem: From Fedora 34 it is mandatory to use valid ip instead of localhost, 127.0.0.1 or loopback address (0.0.0.0 to 0.255.255.255).
Solution: use $hostname -I | awk '{print $1}' instead
Fixes: #2944 Change-Id: I282a0a519c650c6848ffa668ade86a4f35f9de42 Signed-off-by: Shwetha K Acharya [email protected]
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
1 test(s) needed retry ./tests/00-geo-rep/00-georep-verify-setup.t https://build.gluster.org/job/gh_centos7-regression/1809/
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
1 test(s) needed retry ./tests/00-geo-rep/00-georep-verify-setup.t https://build.gluster.org/job/gh_centos7-regression/1818/
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
2 test(s) needed retry ./tests/000-flaky/glusterd-restart-shd-mux.t ./tests/00-geo-rep/00-georep-verify-setup.t https://build.gluster.org/job/gh_centos7-regression/1819/
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
3 test(s) needed retry ./tests/000-flaky/basic_ec_ec-quorum-count-partial-failure.t ./tests/000-flaky/glusterd-restart-shd-mux.t ./tests/00-geo-rep/00-georep-verify-setup.t
1 flaky test(s) marked as success even though they failed ./tests/000-flaky/glusterd-restart-shd-mux.t https://build.gluster.org/job/gh_centos7-regression/1820/
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
2 test(s) needed retry ./tests/000-flaky/glusterd-restart-shd-mux.t ./tests/00-geo-rep/00-georep-verify-setup.t
1 flaky test(s) marked as success even though they failed ./tests/000-flaky/glusterd-restart-shd-mux.t https://build.gluster.org/job/gh_centos7-regression/1821/
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
2 test(s) needed retry ./tests/000-flaky/basic_afr_split-brain-favorite-child-policy.t ./tests/00-geo-rep/00-georep-verify-setup.t https://build.gluster.org/job/gh_centos7-regression/1824/
dependent on: https://github.com/gluster/project-infrastructure/issues/146
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
1 test(s) needed retry ./tests/00-geo-rep/00-georep-verify-setup.t https://build.gluster.org/job/gh_centos7-regression/1833/
1 test(s) failed ./tests/basic/ec/quota.t
0 test(s) generated core
2 test(s) needed retry ./tests/00-geo-rep/georep-basic-dr-rsync.t ./tests/basic/ec/quota.t https://build.gluster.org/job/gh_centos7-regression/1838/
For my own knowledge, can the problem be explained, eg, what changed in Fedora 34 ?
For my own knowledge, can the problem be explained, eg, what changed in Fedora 34 ?
In fedora 34 we cannot use loopback addresses to refer to a host. This is possibly a security / seggregation of loopback address usage, which I am not entirely sure why. @mscherer
On using 127.0.0.1:
[root@localhost glusterfs]# vi tests/00-geo-rep/00-georep-verify-setup.t [root@localhost glusterfs]# prove -v tests/00-geo-rep/00-georep-verify-setup.t tests/00-geo-rep/00-georep-verify-setup.t .. 1..32 ok 1 [ 163/ 1328] < 13> 'glusterd' ok 2 [ 21/ 16] < 14> 'pidof glusterd' Please provide a valid hostname/ip other than localhost, 127.0.0.1 or loopback address (0.0.0.0 to 0.255.255.255).
Usage: volume create <NEW-VOLNAME> [stripe <COUNT>] [replica <COUNT> [arbiter <COUNT>]] [disperse [<COUNT>]] [disperse-data <COUNT>] [redundancy <COUNT>] [transport <tcp|rdma|tcp,rdma>] <NEW-BRICK>... [force]
not ok 3 [ 10/ 63] < 32> 'gluster --mode=script --wignore volume create primary replica 2 127.0.0.1:/d/backends/primary1 127.0.0.1:/d/backends/primary2 127.0.0.1:/d/backends/primary3 127.0.0.1:/d/backends/primary4' -> '' volume start: primary: failed: Volume primary does not exist not ok 4 [ 7/ 74] < 33> 'gluster --mode=script --wignore volume start primary' -> '' Please provide a valid hostname/ip other than localhost, 127.0.0.1 or loopback address (0.0.0.0 to 0.255.255.255).
Usage: volume create <NEW-VOLNAME> [stripe <COUNT>] [replica <COUNT> [arbiter <COUNT>]] [disperse [<COUNT>]] [disperse-data <COUNT>] [redundancy <COUNT>] [transport <tcp|rdma|tcp,rdma>] <NEW-BRICK>... [force]
The commit cc51d2ab9834 is what added the check, from 2019, so that's not a recent change.
I guess something changed in Fedora that make it work.
1 test(s) failed ./tests/basic/glusterd/disperse-create.t
0 test(s) generated core
1 test(s) needed retry ./tests/basic/glusterd/disperse-create.t https://build.gluster.org/job/gh_centos7-regression/1839/
Ok but so:
- why does it still work on EL8 and EL7 ?
- why is the check here in the 1st place if it was broken, and why do we keep it ?
1 test(s) failed ./tests/00-geo-rep/00-georep-verify-setup.t
0 test(s) generated core
2 test(s) needed retry ./tests/000-flaky/glusterd-restart-shd-mux.t ./tests/00-geo-rep/00-georep-verify-setup.t
1 flaky test(s) marked as success even though they failed ./tests/000-flaky/glusterd-restart-shd-mux.t https://build.gluster.org/job/gh_centos7-regression/1872/
- why does it still work on EL8 and EL7 ? We need to investigate, why the check was not functioning in some OSs and OS versions. I will update if I get any more info regarding that. However #2962 existed on all the OSs.
- why is the check here in the 1st place if it was broken, and why do we keep it ? This check is also present in https://github.com/gluster/glusterfs/blob/devel/cli/src/cli-cmd-parser.c#L172 apart from the changes introduced in https://github.com/gluster/glusterfs/commit/cc51d2ab9834f4fc01fcf219bb727d089d5fc271. This check is a generic gluster parsing constraint deployed from the very beginning. Though I am not sure of the exact historical reason, it could be to prevent some functionality of gluster from falling apart.
Context for that change: https://github.com/gluster/glusterfs/commit/930e1882f59b2f36980be9dddda1a3d3f5b8e459
and so https://bugzilla.redhat.com/show_bug.cgi?id=787627