GitLab

This daemon will take over cluster maintenance as per
our design document. As it will heavily depend on the monitoring
daemon, it will only be enabled (at configure time) if the
monitoring daemon is enabled as well. It will also run as
the same user and group.

In this commit only the plain daemon is added with the only supported
request being the question for the supported protocol versions.
Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Petr Pudlak <pudlak@google.com>

This patch modifies the daemon-util to check if setcap is present
before attempting to use it.
Signed-off-by: Hrvoje Ribicic <riba@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

The metadata daemon was previously running as root due to its need to
open port 80 to provide information to instances. To allow the daemon
to run in a more secure way, this patch adds a separate metadata user,
and grants the metad executable the CAP_NET_BIND_SERVICE capability.
As a result, the metadata daemon can use the port 80 without having to
acquire the full set of root capabilities and drop it later.
Signed-off-by: Hrvoje Ribicic <riba@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

The only reason why it was optional was its dependency on regex-pcre,
which was an optional Ganeti dependency.

Job filter rules require rexex-pcre unconditionally, so now confd
requires no additional depdendencies and can be built by default.
Signed-off-by: Niklas Hambuechen <niklash@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

It was decided that Ganeti is relicensed under the 2-clause
BSD license. Update the license statements accordingly (issue #936).
Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Petr Pudlak <pudlak@google.com>

For some reason, find does stat the current working directory,
which, when invoked via cron, may or may not be readable by the
process (issue 880). Therefore, first change to a directory that
is definitely accessible to the process; the directory it logs
into is such a safe place.
Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Petr Pudlak <pudlak@google.com>

Signed-off-by: Petr Pudlak <pudlak@google.com>
Reviewed-by: Jose A. Lopes <jabolopes@google.com>

This patch modifies how the KVM daemon is started/stopped.  After this
patch, the KVM daemon is always started by default, as it will
terminate itself if necessary, and it is restarted by the watcher.

Also, changes to user shutdown and vm capable attributes will cause
the KVM daemon instances on the corresponding nodes to be
started/stopped accordingly.  This is achieved through the helper
function 'EnsureKvmdOnNodes'.
Signed-off-by: Jose A. Lopes <jabolopes@google.com>
Reviewed-by: Hrvoje Ribicic <riba@google.com>

noded is designed to live on a normal node and hence
will not depend on any other daemon. In particular, it
is safe to start this daemon first. Starting it after
wconfd also allows wconfd to carry out a full voting.
Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Petr Pudlak <pudlak@google.com>

... to make lint shut up.
Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Jose Lopes <jabolopes@google.com>

This daemon does not do anything anymore, so we
can as well get rid of it.
Signed-off-by: Klaus Aehlig <aehlig@google.com>
Reviewed-by: Hrvoje Ribicic <riba@google.com>
Signed-off-by: Klaus Aehlig <aehlig@google.com>

We assume that the compression tools the user supplies use stdin and
stdout for handling data, and that a switch is used to distinguish
compression from decompression. This patch introduces these constraints
by adding the invocation of these tools to the import-export daemon.
Signed-off-by: Hrvoje Ribicic <riba@google.com>
Reviewed-by: Thomas Thrainer <thomasth@google.com>

The addition of lzop as a compression option that does not necessarily
exist on a Ganeti-running machine requires checks that lzop exists.
This patch applies a more general solution, and attempts to check for
the existence of a tool by running it with the -h switch on the target
node. A timeout is imposed in case the tool expects input, and missing
tools are handled as well.
Signed-off-by: Hrvoje Ribicic <riba@google.com>
Reviewed-by: Thomas Thrainer <thomasth@google.com>

Ganeti took care to restrict all possible compression invocations to
the few options that were available. This patch strips away all of
those, but does not allow any interesting and dangerous commands
provided to the import-export daemon to be executed yet.
Signed-off-by: Hrvoje Ribicic <riba@google.com>
Reviewed-by: Thomas Thrainer <thomasth@google.com>

daemon-util's check-and-start function is only used by the watcher to
start failed services that should be running on the node. Systemd
supervises and restarts the services as needed; if a service is down, it
is so either because of an unrecoverable error, or because the
administrator manually stopped it. In any case, it should not be blindly
restarted.
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Reviewed-by: Jose A. Lopes <jabolopes@google.com>

daemon-util is used throughout the code for starting a daemon whenever
this is needed (e.g. during cluster bootstrapping or master failover).

In order not to confuse systemd and its service supervision code,
daemon-util needs to cooperate with it. Thus, we use systemctl for all
operations if both of these conditions hold:

 - systemd is running as PID 1, checked via the existence of
   /run/systemd/system, as per sd_booted(3).
 - systemd has the ganeti.target unit loaded, indicating that the
   relevant systemd unit files have been installed.
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Reviewed-by: Jose A. Lopes <jabolopes@google.com>

We add a second category of daemons in daemon-util, those that are
started on demand and should not be started by the start_all action, but
should be stopped during service shutdown. This category currently
includes metad and kvmd.

To this end we introduce two more array variables:

 - ON_DEMAND_DAEMONS: currently holds kvmd and metad, which should only
   be stopped during stop-all.

 - ALL_DAEMONS: is the union of DAEMONS and ON_DEMAND_DAEMONS, i.e. the
   list of all known daemons.

We also change the daemon-util test to expect kvmd and metad in the stop
daemon list.
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Reviewed-by: Jose A. Lopes <jabolopes@google.com>

* Add 'ganeti-metad' to list of daemons to stop
* Update 'daemon-util' test

Note that 'ganeti-metad' is not automatically started together with
all the other daemons.  Instead, 'ganeti-metad' is only started when
it is necessary.  Nevertheless, it is stopped together with the other
daemons.
Signed-off-by: Jose A. Lopes <jabolopes@google.com>
Reviewed-by: Hrvoje Ribicic <riba@google.com>

.. and make it the first one so that masterd can use it when it starts.
Signed-off-by: Petr Pudlak <pudlak@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

The daemon exposes the declared functions in Ganeti.WConfd.Core to RPC
clients (currently just 'echo').
Signed-off-by: Petr Pudlak <pudlak@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

and update the unit check the list of daemons.
Signed-off-by: Petr Pudlak <pudlak@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

The Node deamon was executed with the wrong gid (gnt-daemons) instead of the one
assigned to it by configure.ac.

Fixes Issue 707.
Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

In 'daemons/daemon-util.in', 'start-master' and 'stop-master' always
fail if confd is disabled.

Fixes issue 685.
Signed-off-by: Jose A. Lopes <jabolopes@gmail.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

Luxid was not handled in start_master() and stop_master() at all. As a result,
during a master-failover, luxid would be left running on the old master and
would not start on the new master, leaving the cluster without management until
luxid was manually started.
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Michele Tartara <mtartara@google.com>

daemon-util's rotate_logs() did not pass --oknodo to start-stop-daemon while
HUPing the daemon processes. As a result, rotate_logs would fail for a
non-running daemon causing rotate_all_logs to exit prematurely.

We can fix this by either passing --oknodo to start-stop-daemon, or have
rotate_all_logs ignore any errors. The former is probably a safer approach - as
it doesn't mask any potential coding errors in rotate_logs - and is also
semantically correct (we only care about log rotation if a daemon is actually
running).
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Modify daemon-util to allow sending SIGHUP to one or all daemons. This is meant
as a utility function to be used in logrotate definitions.
Signed-off-by: Apollon Oikonomopoulos <apoikos@gmail.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Commit 91525dee fixed Issue 477 but broke
"gnt-cluster info".

This commit offers a solution to both problems, by changing the permission
of the socket instead of changing the permission the confd process is run
with.
Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

As queryd will, in the future, handle all LUXI request, queue jobs and
most likely perform various other tasks, it is renamed to luxid already.
This will safe some headache when upgrading Ganeti installations, as we
don't have to deal with a daemon rename.
Signed-off-by: Thomas Thrainer <thomasth@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

queryd is added as a new daemon which handles configuration queries over
LUXI. This functionality was removed from confd, which now only queries
over the network.

The queryd user is added to the master group such that it can access
server.pem. The confd user no longer needs to be in the master group.
This fixes issue 292.
Signed-off-by: Thomas Thrainer <thomasth@google.com>
Reviewed-by: Helga Velroyen <helgav@google.com>

Starting confd as a member of the daemons group allows the RAPI daemon to access
the LUXI socket.

Fixes Issue 477.
Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Klaus Aehlig <aehlig@google.com>

This is a workaround for issue 477. Confd resets the
permissions of the query socket in a wrong way. This
patch fixes them after the start of confd.
Signed-off-by: Helga Velroyen <helgav@google.com>
Reviewed-by: Thomas Thrainer <thomasth@google.com>

This commit adds the core infrastructure of the monitoring daemon,
and integrates it in the build and test systems.

The actual functionality of the monitoring daemon is still completely
missing.
Signed-off-by: Michele Tartara <mtartara@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Signed-off-by: Dimitris Aragiorgis <dimara@grnet.gr>
Reviewed-by: Iustin Pop <iustin@google.com>

As I wrote during/after the review on commit 2958c56e, “ganeti-cleaner:
Separate queue cleaning code”, while I appreciated the permission
separation, I didn't like too much the file-based approach:

- it is a very simple script, and lots of the code is duplicated
  between the two; I wouldn't like to see "ganeti-vmcapable-cleaner",
  "ganeti-master-candidate-cleaner", etc. in the future
- ganeti-master-cleaner "pollutes" the namespace, creating
  tab-completion conflicts with ganeti-masterd

This patch simply merges the master-cleaner back into cleaner, while
keeping the separate user permissions scheme, separate log files, etc.

Additionally, it fixes two bugs in the unit-test (not run with set -u
and wrong path in the master-cleaner log files test; yay for even
worse safety than Python?).

And finally, since we have now support for --help-completion, it adds
bash completion support for this script :) (needs to be applied on top
of my argument support patch series).
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

This code does not need to run as root, therefore it's better to split
it out. It is now run with the same permissions as the master daemon.
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Instead of having a different set in (almost) every shell script, this
inserts the most commonly used variables at build time. This way the
code for injecting a root directory for virtual clusters also is just
needed once.

I chose inserting the variables using “sed” at build time over a runtime
script as the latter would need to be located from within the script and
the path would be different depending on the environment (unittest vs.
actual installation).
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

… instead of comparing with two different values in two placse.
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

GANETI_ROOTDIR contains the root directory for the current “virtual node”.
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

Commit 4b42c3d6 broke non-root user mode since, while trying to do a
cleanup a move all local variable definitions to the start of the
function; however, the plain_name var is only defined later, so this
actually doesn't work.

Note that enabling set -u doesn't _directly_ work, since we rely on
some variables which are usually undefined (EXTRA_*_ARGS, etc.). It
can be fixed, but in a later patch.
Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: René Nussbaumer <rn@google.com>

This adds support to daemon-util for Red Hat based distributions that
do not have a start-stop-daemon. If /sbin/start-stop-daemon is not
available, daemon-util will source /etc/rc.d/init.d/functions.
check(), start(), and stop() are updated to use the relevant functions
from /etc/rc.d/init.d/functions.

Thanks to Stephen Fromm and Michael Hanselmann for improving the error
handling, style, and comments.
Signed-off-by: Jun Futagawa <jfut@integ.jp>
Reviewed-by: Michael Hanselmann <hansmi@google.com>
(cherry picked from commit 4b42c3d6)