With this patch all timeouts are pre-calculated. The interface of
the _LockTimeoutStrategy class is also changed a bit; NextAttempt
now returns a new instance.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

The timeout is always between ~0.1 and ~10.0 seconds. A small
variation of ±5% is added to prevent different jobs from
fighting each other. After 10 attempts to acquire the locks with
a timeout, a blocking acquire is made.

Lock status reporting will be improved in a separate patch.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

The timeout passed to LockSet.acquire() is measured over all lock acquires. If
LockSet.acquire fails to acquire all requested locks within the specified
amount of time, all locks are released again and the acquire fails.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

- We never remember to use it (5 uses vs 21 " ,".join())
- It's longer to write than " ,".join()
- The added value of the apostrophe in the string is not very much

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Now featuring unit testing, and more deterministic results on some
corner cases.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

There were more race conditions. By adding a notify function to
SharedLock.acquire we can prevent them.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

With this patch we simplify usage on the SingleActionCondition (which
wasn't a condition at all) by making it a real condition. This way we
can just wait() on it, or notifyAll() as we would on a normal one. The
only catch is that notifyAll can be called only once, and wait can only
be called before notifyAll has, but luckily our PipeCondition, now quite
simplified, takes care of this, by providing a new SingleActionCondition
each time the previous one has been notified.

No Start/StopUsing function are needed anymore, and thus the condition
is a lot more robust, and there's no way file descriptors can be left
open, as they are closed in a finally block, in the same function where
they were opened, by the last thread exiting the class.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

This way they can be used to test different condition classes.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

All (ok, all but one) _ThreadedTestCase users have a done Queue, so we
move its building in the _ThreadedTestCase setUp

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

SharedLock.__acquire_unlocked uses keyword parameters. Just passing
the timeout would set the “shared” parameter.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Also remove the “blocking” parameter from LockSet.remove and
GanetiLockManager.remove. There's no point in implementing timeouts on removal
unless we need them.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Our automated test system found a few problems in the new locking
unittests. This patch should fix them, although I wasn't able to
reproduce the problem. All are race conditions.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

_PipeCondition is a condition implemented using pipe(2) and poll(2).
It allows the implementation of timeouts without using a busy-wait loop
with time.sleep.

Unlike Python's built-in threading.Condition class and to save file
descriptors and an internal queue, it can only be used to notify
all waiters. Ganeti's use case for this condition class doesn't
require the ability to notify only one waiter.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This class will be used as a basic block for pipe(2)-based
conditions. Upon initialization it creates a pipe and can be
notified once (hence the “single action” in the name). A
callable helper class is used to wait for notifications.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

This patch greatly simplifies the SharedLock code and implements
timeouts for the acquire() and delete() functions. A wrapper around
Python's threading.Condition class must be used to ensure thread
safety when check whether there are any waiters left.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Some basic tests for the confd client library

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

While ‘test1’ matches both ‘test1’ and ‘test1.example’, it has a full,
exact match and we should return it if that is the case.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Checking that the fourcc code is indeed a fourcc code, and that request
codes and reply statuses are distinct.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Luca Bigliardi <shammash@google.com>

It should only be used at build-time.

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Luca Bigliardi <shammash@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Luca Bigliardi <shammash@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Luca Bigliardi <shammash@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

The mcpu.py:HooksMaster class needs to have a proc attribute/argument to
init in ordet to call its LogWarning method. However, this is available
from the 'lu' attribute, so we can remove this dependency.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

In case there are gaps in /proc/drbd related to the minors sequence,
this can lead to empty lines (see the attached data file). In this case,
we currently fail to parse the file correctly.

The patch fixes this by skipping empty lines in
bdev.BaseDRBD._MassageProcData and adds a data file and unittest for
this condition.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

We don't format with subsecond-precision.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

The salt needs to be returned anyway, so we don't have to add another
key for the sender to recognize which request an answer is answering, so
all that infrastructure is useless. :(

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

The method is changed to a normal loop, to avoid calling getattr()
twice. Also __getstate__ is changed to just use ToDict() by default.

This should also make __getstate__ work for objects which have to
override the ToDict function because they contain other objects.

__setstate__ is probably still broken in this case, but so it was
before, and it's not used inside our code, so I'll pretend not to have
noticed, as there is no "nice" way to fix it, without overriding it all
over the place :(

Some unittests are added as a bonus, to make sure we behave well.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

File "../test/ganeti.hooks_unittest.py", line 239, in setUp
  self.lu = FakeLU(FakeProc(), self.op, self.context, None)
File "…/ganeti/cmdlib.py", line 92, in __init__
  self.LogStep = processor.LogStep
AttributeError: FakeProc instance has no attribute 'LogStep'

Signed-off-by: Michael Hanselmann <hansmi@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This parameter is now mandatory for the cluster config to work.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

Currently we have both a default_hypervisor and an enabled_hypervisors
list. The former is only settable at cluster init time, while the latter
can be changed with cluster modify.

This becomes cumbersome in a few ways: at cluster init time for example
if we pass in a list of enabled hypervisors which doesn't include the
"default" xen-pvm one, we're also forced to pass a default hypervisor,
or an error will be reported. It is also currently possible to disable
the default hypervisor in cluster-modify (with unknown results).

In order to avoid this we get rid of this field altogether, and define
the "first" enabled hypervisor as the default one. This allows ease of
changing which one is the default, and at the same time maintains
coherency.

At configuration upgrade we make sure that the old default is first in
the list, so that 2.0 cluster defaults are preserved.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

This patch includes HMAC authenticated json messages to the serializer.
The new interface works on any json-encodable data type, and can sign it
with a private key and an optional salt. The same private key must be
used upon message loading to verify the message.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Michael Hanselmann <hansmi@google.com>

If an ident member of an IdentKeyVal relationship starts with no_ or -,
handle it the same way we do for a key. Some unittests are added to
check that check_ident_key_val behaves as expected.

This patch also changes ForceDictType to, for now, fail on such an
entry, and the same to happen when creating an instance or modifying its
nics or disks.

This behavior will be used later on to allow deletion of os entries in
os parameters.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

If an empty string is passed to _SplitKeyVal, we should return {},
rather than {'': True}. Also test for the correct behavior.

Signed-off-by: Guido Trotter <ultrotter@google.com>
Reviewed-by: Iustin Pop <iustin@google.com>

Patch "Simplify the RPC result framework in backend.py" changed all
RPCs, and thus the hooks results where also changed. This needs changes
to the hooks unittests too.

Signed-off-by: Iustin Pop <iustin@google.com>
Reviewed-by: Guido Trotter <ultrotter@google.com>