Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 08 Jul 2019 16:39:36 +0000
From:      bugzilla-noreply@freebsd.org
To:        ports-bugs@FreeBSD.org
Subject:   [Bug 238785] sysutils/py-google-compute-engine: Some GCE packages fill the logs with an error message
Message-ID:  <bug-238785-7788-9fqW7w8tZn@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-238785-7788@https.bugs.freebsd.org/bugzilla/>
References:  <bug-238785-7788@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D238785

--- Comment #8 from Mage <mage@mage.gold> ---
I did my best. The rest is up to you. Here is what I found:


I created two servers, test1 and test2. test1 is a clone or a running serve=
r.
test2 was a saved image of that server in January 2019.

I booted up both servers. test1 put the error message into the log every
minute. test2 did not put there anything.

I upgraded both servers using pkg upgrade. I compared the package list with
diff. They were identical except for the PostgreSQL version (10 vs. 11). It=
 was
the only difference.

test1 kept adding the error to the logs. test2 did not do it.

(All my running servers behave the same as test1).

I ran diff -ur on /usr/local/etc and /etc to compare the both servers. I fo=
und
almost nothing belpful. They were almost the same, except fot the apache co=
nfig
in /usr/local/etc.

There was in interesting difference though. On test2, my user was in the
google-sudoers group. On test1, it was not.

google-sudoers:*:1003:
vs.=20
google-sudoers:*:1003:mage

I don=E2=80=99t like sudo. On my non-google FreeBSD servers, I=E2=80=99ve n=
ever had it
installed. On GCE, I use it solely for 'sudo su -'

It is sure I did not remove myself from that file. I didn=E2=80=99t even kn=
ow I was in
that group. I don=E2=80=99t even know what that group does because I can 's=
udo su -' on
any server on which I=E2=80=99m not a member of that group.

As the log was not helpful, I went throgh the three files, and I added a li=
ne
to /usr/local/lib/python2.7/subprocess.py

Line 185:
print popenargs

Then, I killed the daemons (the service google-xxx-daemon stop doesn=E2=80=
=99t work by
the way), and I ran it from console.

root@test1 ~ # /usr/local/bin/google_accounts_daemon-2.7
(['gpasswd', '-a', 'mage', 'google-sudoers'],)
(['gpasswd', '-a', 'mage', 'google-sudoers'],)
[...]

It tries to add me to that group in every minute using gpasswd. There is no
gpasswd on the server, and I think it never was.

For this reason, test2 also couldn=E2=80=99t have gpasswd. I was surprised =
because it
did the following:

root@test2 ~ # /usr/local/bin/google_accounts_daemon-2.7
(['pw', 'groupmod', 'google-sudoers', '-m', 'mage'],)

For some reason, test2 didn=E2=80=99t try to execute gpasswd (Linux) but th=
e pw
command, as it should on FreeBSD.

Another difference is that test2 executed the command only once, right afte=
r I
started the daemon. It never tried it again. test1 tried to do it in every
minute but with a different command.

I added myself to that group on test1.

pw usermod mage -G google-sudoers

It did not help. The daemon was trying to add me, with the wrong command.

Now more or less the only difference between the two servers is the FreeBSD
version.

FreeBSD test1 11.2-RELEASE-p9 FreeBSD 11.2-RELEASE-p9 #0: Tue Feb  5 15:30:=
36
UTC 2019     root@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERI=
C=20
amd64

FreeBSD test2 11.2-RELEASE-p8 FreeBSD 11.2-RELEASE-p8 #0: Tue Jan  8 21:35:=
12
UTC 2019     root@amd64-builder.daemonology.net:/usr/obj/usr/src/sys/GENERI=
C=20
amd64

I=E2=80=99m not sure this causes them to run different commands for adding =
me to the
group. I will run freebsd-update on both, and see what happens.

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-238785-7788-9fqW7w8tZn>