Mailing-List: contact user-help@zookeeper.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@zookeeper.apache.org
Date: Wed, 15 Jul 2015 13:19:30 -0500
From: Jordan Zimmerman <jordan@jordanzimmerman.com>
To: zookeeper-user@hadoop.apache.org, Vikas Mehta
 <vikasmehta@gmail.com>, user@zookeeper.apache.org
Message-ID: <etPan.55a6a432.3ca70d1c.15a@Jordans-MacBook-Pro.local>
In-Reply-To: <1436984221201-7581279.post@n2.nabble.com>
References: <1436982861611-7581277.post@n2.nabble.com>
 <CABWqe2YKsbBtUb93FEdd23ffOJWQc+wbbwhZbVEeME9b9z-=Rg@mail.gmail.com>
 <1436984221201-7581279.post@n2.nabble.com>
Subject: Re: locking/leader election and dealing with session loss
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="55a6a432_7c111597_15a"

--55a6a432_7c111597_15a
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

Once client A loses connection it must assume that it no longer has the l=
ock (you could try to time the session but I think that=E2=80=99s a bad i=
dea). Once you reconnect, you will know if your session is still active o=
r not. When done correctly, there=E2=80=99s no chance that both A and B w=
ill think they own the lock at the same time.

-Jordan


On July 15, 2015 at 1:17:10 PM, Vikas Mehta (vikasmehta=40gmail.com) wrot=
e:

Thanks for the quick response Camille. If client A owns the lock, gets =20
disconnected due to network partition, it will not see the SESSION=5FEXPI=
RED =20
event until it is too late, i.e. client B has acquired the lock and done =
the =20
damage. Problem here is that client cannot distinguish network partition =
=20
from zookeeper ensemble in leader election state. =20


-- =20
View this message in context: http://zookeeper-user.578899.n2.nabble.com/=
locking-leader-election-and-dealing-with-session-loss-tp7581277p7581279.h=
tml =20
Sent from the zookeeper-user mailing list archive at Nabble.com. =20

--55a6a432_7c111597_15a--