[ovirt-devel] URGENT! OST broken! please stop merging patches to ovirt-engine!

[Barak Korren]

We have two unresolved issues that are currently causing all master
OST runs to fail and preventing us from effectively finding
regressions.

The 1st issue, which was already reported on Sunday, is a regression
that is causing vm_run to fail and was probably introduced by this
patch:
https://gerrit.ovirt.org/#/c/79033/41

Here is a recent run that failed with this:
http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/2147/

Logs are here:
http://jenkins.ovirt.org/job/ovirt-master_change-queue-tester/2147/artifact/exported-artifacts/basic-suit-master-el7/test_logs/basic-suite-master/post-002_bootstrap.py/

The 2nd issue is a new one, that seems to be causing add-host failures
and I've just send another email about it. The gist of this is that
I'm siing the following failures in the supervdsm logs, and vdsm is
not loading:

MainThread::ERROR::2017-08-30
05:55:59,476::initializer::53::root::(_lldp_init) Failed to enable
LLDP on eth1
Traceback (most recent call last):
  File "/usr/lib/python2.7/site-packages/vdsm/network/initializer.py",
line 51, in _lldp_init
    Lldp.enable_lldp_on_iface(device)
  File "/usr/lib/python2.7/site-packages/vdsm/network/lldp/lldpad.py",
line 30, in enable_lldp_on_iface
    lldptool.enable_lldp_on_iface(iface, rx_only)
  File "/usr/lib/python2.7/site-packages/vdsm/network/lldpad/lldptool.py",
line 46, in enable_lldp_on_iface
    raise EnableLldpError(rc, out, err, iface)
EnableLldpError: (1,
"timeout\n'M00000001C3040000000c04eth1000badminStatus0002rx' command
timed out.\n", '', 'eth1')
MainThread::DEBUG::2017-08-30
05:55:59,477::cmdutils::133::root::(exec_cmd) /usr/sbin/lldptool
get-lldp -i eth0 adminStatus (cwd None)


This failure does not seem to have been introduced by the platform and
not oVirt code, because it also happens on code that already passed
OST a few days ago.

Please avoid merging any patches except to the purpose of resolving
these issues. Nothing is making it into the 'tested' and nightly
snapshot repos anyway ATM.

-- 
Barak Korren
RHV DevOps team , RHCE, RHCi
Red Hat EMEA
redhat.com | TRIED. TESTED. TRUSTED. | redhat.com/trusted