qpid-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From girish lc <girish...@gmail.com>
Subject Re: Qpid High Availability
Date Fri, 14 Jul 2017 17:54:42 GMT
Hi Alex,

Thank you,

After set-up of the 3 node cluster (I,e Quorum nodes), I can see cluster
continues to operate after losing one node. Also I verified the message is
reached to the new MASTER node.


On Fri, Jul 14, 2017 at 5:40 PM, Oleksandr Rudyy <orudyy@gmail.com> wrote:

> Hi Girish,
>
> HA cluster operates when the majority of nodes is available (when "Required
> Minimum Number Of Nodes" is set to DEFAULT).
> In 2-node cluster the majority is 2. Thus, when one node goes down, the
> cluster looses majority and cannot operate even when second node is Ok.
> It is implemented like this in order to avoid "Split brain scenario".
>
> You can make cluster operational by setting "Allow to Operate Solo" on the
> remaining node.
>
> Please note, that only one node in 2-node cluster can have "Allow to
> Operate Solo" set on.
>
> Alternately, you can set-up 3-node cluster. The 3-nodes cluster continues
> to operate after loosing one node (due to majority being 2, thus, the
> election of the Master can be performed).
> Obviously, when 2 nodes are lost in 3-nodes cluster, the remaining node
> cannot operate by itself (unless you override "Required Minimum Number Of
> Nodes" and set it to 1).
>
>
> I hope that the above can help you to understand why node2 goes into
> WAITING Role and node1 becomes UNREACHABLE on stop of virtual machine for
> node1.
>
> Kind Regards,
> Alex
>
>
>
> On 14 July 2017 at 11:41, girish lc <girish.lc@gmail.com> wrote:
>
> >
> >
> > Hi Alex,
> >
> > Thank you very much for your detailed email, now I'm able to send the
> > message in the group.
> >
> > I had used the below URL with fail over settings,
> >
> > connectionfactory.qpidConnectionFactory=amqp://guest:guest@clientid
> > /hagroup?failover='roundrobin?cyclecount='2''&brokerlist='
> > tcp://<node1>:5672?retries='5'&connectdelay='2000';tcp://<
> > node2>:5672?retries='5'&connectdelay='2000''
> >
> > 1. When I do the flip b/w MASTER and REPLICA, I can see the messages in
> > the new MASTER(old REPLICA) .
> >
> > 2. But when I stop the VM1(node1) system the VM2(node2) goes to WAITING
> > Role and node1 becomes UNREACHABLE and the message sent to MASTER will
> not
> > deliver and gives the below error,
> > Could you please let me know how to handle the Failover case of MASTER
> > node OR Is it the correct behavior ???
> >
> >
> > [image: Inline image 4]
> >
> >
> >
> >
> > [image: Inline image 3]
> >
> >
> >
> >
> >
> > On Thu, Jul 13, 2017 at 7:27 PM, Oleksandr Rudyy <orudyy@gmail.com>
> wrote:
> >
> >> Hi Girish,
> >>
> >> As a simple test you can publish a message into your HA virtual host,
> flip
> >> the mastersheep to replica, and consume that message from a new master.
> >>
> >> The connection URL you have provided is from JMS client for AMQP 0-x
> >> protocols [1]. You can use JMS client for AMQP 0-x or switch to newer
> jms
> >> client for AMQP 1.0 [2].
> >>
> >> For both JMS client you need to specify a virtual host name in a
> >> connection
> >> URL if ha node is not a default one.
> >>
> >> For example, AMQP 0-x client connection URL would need to be changed as
> >> below
> >>
> >> connectionfactory.qpidConnectionFactory=amqp://guest:guest@clientid/<ha
> >> virtual host name>?brokerlist='tcp://localhost:5672'
> >>
> >> where <ha virtual host name> is your HA group name.
> >>
> >> The format of connection URL for AMQP 0-x client is documented in [2]:
> >>
> >> If you need an automatic failover (client reconnects automatically to
> >> master broker on mastersheep change), you can specify both brokers in a
> >> connection URL with failover settings,  for example
> >>
> >> connectionfactory.qpidConnectionFactory=amqp://guest:guest@/<ha virtual
> >> host name>?failover='roundrobin?cyclecount='2''
> >> &brokerlist='tcp://<host1>:<port1>?retries='5'&connectdelay=
> >> '2000';tcp://<host2>:<port2>?retries='5'&connectdelay='2000''
> >>
> >> For new JMS client please check documentation at [4].
> >>
> >> Kind Regards,
> >> Alex
> >>
> >> [1] http://qpid.apache.org/components/jms/amqp-0-x.html
> >> [2] http://qpid.apache.org/components/jms/index.html
> >> [3]
> >> http://qpid.apache.org/releases/qpid-java-6.1.4/jms-client-
> >> 0-8/book/JMS-Client-0-8-Connection-URL.html
> >> [4] https://qpid.apache.org/releases/qpid-jms-0.23.0/docs/index.html
> >>
> >>
> >>
> >> On 13 July 2017 at 14:36, girish lc <girish.lc@gmail.com> wrote:
> >>
> >> > Hi,
> >> >
> >> > I'm able to create group as mentioned in the link,
> >> > [image: Inline image 1]
> >> > http://qpid.apache.org/releases/qpid-java-6.1.4/java-
> >> > broker/book/Java-Broker-High-Availability-CreatingGroup.html
> >> >
> >> > Somebody please guide me how to test the HA.
> >> >
> >> > Here below are the steps I followed:
> >> >
> >> > 1. Created the MASTER node in one VM(1st VM). and added the other
> VM(2nd
> >> > VM) details in the permitted address.
> >> >
> >> > 2. Created the REPLICA in the 2nd VM and added MASTER node details in
> >> the
> >> > permitted address
> >> >
> >> > 3. When I was sending the message to the Queue earlier I was using
> this
> >> > below line in the config.properties file.
> >> >
> >> > connectionfactory.qpidConnectionFactory=amqp://guest:guest@clientid/?
> >> > brokerlist='tcp://localhost:5672'
> >> >
> >> > But now I am not able to send any message to the MASTER node, getting
> >> the
> >> > below error.
> >> >
> >> > [image: Inline image 2]
> >> >
> >> >
> >> >
> >> >
> >> >
> >> > On Wed, Jul 12, 2017 at 11:13 AM, girish lc <girish.lc@gmail.com>
> >> wrote:
> >> >
> >> >> Hi Alex,
> >> >>
> >> >> I'm following the same link, now I've created one MASTER node and one
> >> >> REPLICA to form the group, but I am not getting the idea how to use
> or
> >> test?
> >> >>
> >> >> On Tue, Jul 11, 2017 at 6:25 PM, Oleksandr Rudyy <orudyy@gmail.com>
> >> >> wrote:
> >> >>
> >> >>> Girish,
> >> >>>
> >> >>> Please following instructions in Qpid documentation [1].
> >> >>>
> >> >>> Kind Regards,
> >> >>> Alex
> >> >>>
> >> >>> [1]
> >> >>> http://qpid.apache.org/releases/qpid-java-6.1.4/java-broker/
> >> >>> book/Java-Broker-High-Availability-CreatingGroup.html
> >> >>>
> >> >>> On 11 July 2017 at 11:10, girish lc <girish.lc@gmail.com>
wrote:
> >> >>>
> >> >>> > Hi ,
> >> >>> >
> >> >>> > After changing the port to 5674 (new port) I'm able to create
> Master
> >> >>> node
> >> >>> > :)
> >> >>> >
> >> >>> > Working on the REPLICA node, still if any of you have the
snapshot
> >> for
> >> >>> the
> >> >>> > same please share, I am not sure how to check the High
> Availability
> >> >>> > working, please guide me.
> >> >>> >
> >> >>> > On Tue, Jul 11, 2017 at 3:33 PM, girish lc <girish.lc@gmail.com>
> >> >>> wrote:
> >> >>> >
> >> >>> >> Hi Adel,
> >> >>> >>
> >> >>> >> After including the bdb je 5.0.104, I'm still getting
the below
> >> >>> error, if
> >> >>> >> any of you have already configured could you please send
me the
> >> >>> screen shot.
> >> >>> >>
> >> >>> >> [image: Inline image 1]
> >> >>> >>
> >> >>> >> On Tue, Jul 11, 2017 at 3:03 PM, Adel Boutros <
> >> Adelboutros@live.com>
> >> >>> >> wrote:
> >> >>> >>
> >> >>> >>> Hello,
> >> >>> >>>
> >> >>> >>>
> >> >>> >>> You can find it on Oracle's website:
> >> >>> >>>
> >> >>> >>> http://www.oracle.com/technetwork/database/database-technolo
> >> >>> >>> gies/berkeleydb/downloads/index-098622.html
> >> >>> >>>
> >> >>> >>>
> >> >>> >>> Regards,
> >> >>> >>>
> >> >>> >>> Adel
> >> >>> >>>
> >> >>> >>>
> >> >>> >>> ________________________________
> >> >>> >>> From: girish lc <girish.lc@gmail.com>
> >> >>> >>> Sent: Tuesday, July 11, 2017 11:04:31 AM
> >> >>> >>> To: users@qpid.apache.org
> >> >>> >>> Subject: Re: Qpid High Availability
> >> >>> >>>
> >> >>> >>> Hi Alex,
> >> >>> >>>
> >> >>> >>> Could you please share me the location from where
I can download
> >> bdb
> >> >>> je
> >> >>> >>> 5.0.104, as I'm not able to find.
> >> >>> >>>
> >> >>> >>> Many thanks,
> >> >>> >>>
> >> >>> >>> On Tue, Jul 11, 2017 at 1:40 PM, Oleksandr Rudyy <
> >> orudyy@gmail.com>
> >> >>> >>> wrote:
> >> >>> >>>
> >> >>> >>> > Hi Girish,
> >> >>> >>> >
> >> >>> >>> > I tested HA with bdb je 5.0.73 and got the same
error. It
> >> appears
> >> >>> that
> >> >>> >>> je
> >> >>> >>> > configuration setting "je.cleaner.adjustUtilization"
was
> >> >>> introduced in
> >> >>> >>> > versions after 5.0.73. Unfortunately, the setting
is
> hard-coded
> >> in
> >> >>> the
> >> >>> >>> Qpid
> >> >>> >>> > code. In order to resolve the issue you need
to upgrade bdb je
> >> to
> >> >>> >>> 5.0.104.
> >> >>> >>> >
> >> >>> >>> > Kind Regards,
> >> >>> >>> > Alex
> >> >>> >>> >
> >> >>> >>> > On 10 July 2017 at 16:50, girish lc <girish.lc@gmail.com>
> >> wrote:
> >> >>> >>> >
> >> >>> >>> > > Hi Alex,
> >> >>> >>> > >
> >> >>> >>> > > Thank you for the email,
> >> >>> >>> > >
> >> >>> >>> > > I'm using the below version:
> >> >>> >>> > >
> >> >>> >>> > > qpid-java-6.1.4
> >> >>> >>> > > Berkeley DB Java Edition - 5.0.73
> >> >>> >>> > >
> >> >>> >>> > > Please let me know the correct combination.
> >> >>> >>> > >
> >> >>> >>> > > On Mon, Jul 10, 2017 at 6:54 PM, Oleksandr
Rudyy <
> >> >>> orudyy@gmail.com>
> >> >>> >>> > wrote:
> >> >>> >>> > >
> >> >>> >>> > > > Hi Girish,
> >> >>> >>> > > >
> >> >>> >>> > > > What version of bdb je library are
you using?
> >> >>> >>> > > >
> >> >>> >>> > > > 6.x versions of Qpid Broker are based
on bdb je version
> >> 5.0.104
> >> >>> >>> due to
> >> >>> >>> > > > Oracle license change.
> >> >>> >>> > > >
> >> >>> >>> > > > Master and new 7.x versions of the
Broker will be based
> new
> >> >>> bdb je
> >> >>> >>> > > versions
> >> >>> >>> > > > ( 7.3.7 and above), as Oracle started
to distribute je
> with
> >> >>> dual
> >> >>> >>> > > licenses -
> >> >>> >>> > > > one of which is Apache license.
> >> >>> >>> > > >
> >> >>> >>> > > > I am guessing that you are trying to
start 6.x version of
> >> >>> Broker
> >> >>> >>> with
> >> >>> >>> > bd
> >> >>> >>> > > je
> >> >>> >>> > > > 7.3.x.  Please, use je version 5.0.104.
> >> >>> >>> > > >
> >> >>> >>> > > > Kind Regards,
> >> >>> >>> > > > Alex
> >> >>> >>> > > >
> >> >>> >>> > > >
> >> >>> >>> > > > On 10 July 2017 at 14:03, Chandrashekar,
Girish <
> >> >>> >>> > > gchandrashekar@wabtec.com
> >> >>> >>> > > > >
> >> >>> >>> > > > wrote:
> >> >>> >>> > > >
> >> >>> >>> > > > > Hi Team,
> >> >>> >>> > > > >
> >> >>> >>> > > > >
> >> >>> >>> > > > > Could you please help me to setup
Qpid High Availability
> >> >>> >>> > functionality.
> >> >>> >>> > > > >
> >> >>> >>> > > > >
> >> >>> >>> > > > > I am referring the below link,
but I am not successful
> to
> >> >>> setup;
> >> >>> >>> in
> >> >>> >>> > > fact
> >> >>> >>> > > > I
> >> >>> >>> > > > > got the "422 - je.cleaner.adjustUtilization
is not a
> valid
> >> >>> BDBJE
> >> >>> >>> > > > > environment parameter" error 
and also I had lost the
> >> files
> >> >>> and
> >> >>> >>> > folders
> >> >>> >>> > > > > present in my C:\ after this error.
I tried the same in
> 2
> >> >>> virtual
> >> >>> >>> > > > machines
> >> >>> >>> > > > > and also in my local machine same
result 😞
> >> >>> >>> > > > >
> >> >>> >>> > > > >
> >> >>> >>> > > > > https://qpid.apache.org/releases/qpid-java-6.1.4/java-
> >> >>> >>> > > > > broker/book/Java-Broker-High-A
> >> vailability-CreatingGroup.html
> >> >>> >>> > > > >
> >> >>> >>> > > > >
> >> >>> >>> > > > > Can someone let me know the root
cause and also please
> >> help
> >> >>> me to
> >> >>> >>> > setup
> >> >>> >>> > > > > the High Availability functionality.
> >> >>> >>> > > > >
> >> >>> >>> > > > >
> >> >>> >>> > > > > Regards,
> >> >>> >>> > > > >
> >> >>> >>> > > > > Girish LC
> >> >>> >>> > > > >
> >> >>> >>> > > > > 09886521716
> >> >>> >>> > > > >
> >> >>> >>> > > > > This email and any attachments
are only for use by the
> >> >>> intended
> >> >>> >>> > > > > recipient(s) and may contain legally
privileged,
> >> >>> confidential,
> >> >>> >>> > > > proprietary
> >> >>> >>> > > > > or otherwise private information.
Any unauthorized use,
> >> >>> >>> reproduction,
> >> >>> >>> > > > > dissemination, distribution or
other disclosure of the
> >> >>> contents
> >> >>> >>> of
> >> >>> >>> > this
> >> >>> >>> > > > > e-mail or its attachments is strictly
prohibited. If you
> >> have
> >> >>> >>> > received
> >> >>> >>> > > > this
> >> >>> >>> > > > > email in error, please notify
the sender immediately and
> >> >>> delete
> >> >>> >>> the
> >> >>> >>> > > > > original. Neither this information
block, the typed name
> >> of
> >> >>> the
> >> >>> >>> > sender,
> >> >>> >>> > > > nor
> >> >>> >>> > > > > anything else in this message
is intended to constitute
> an
> >> >>> >>> electronic
> >> >>> >>> > > > > signature unless a specific statement
to the contrary is
> >> >>> >>> included in
> >> >>> >>> > > this
> >> >>> >>> > > > > message.
> >> >>> >>> > > > >
> >> >>> >>> > > >
> >> >>> >>> > >
> >> >>> >>> > >
> >> >>> >>> > >
> >> >>> >>> > > --
> >> >>> >>> > > Regards,
> >> >>> >>> > > Girish LC
> >> >>> >>> > > #09886521716
> >> >>> >>> > >
> >> >>> >>> >
> >> >>> >>>
> >> >>> >>>
> >> >>> >>>
> >> >>> >>> --
> >> >>> >>> Regards,
> >> >>> >>> Girish LC
> >> >>> >>> #09886521716
> >> >>> >>>
> >> >>> >>
> >> >>> >>
> >> >>> >>
> >> >>> >> --
> >> >>> >> Regards,
> >> >>> >> Girish LC
> >> >>> >> #09886521716
> >> >>> >>
> >> >>> >
> >> >>> >
> >> >>> >
> >> >>> > --
> >> >>> > Regards,
> >> >>> > Girish LC
> >> >>> > #09886521716
> >> >>> >
> >> >>>
> >> >>
> >> >>
> >> >>
> >> >> --
> >> >> Regards,
> >> >> Girish LC
> >> >> #09886521716
> >> >>
> >> >
> >> >
> >> >
> >> > --
> >> > Regards,
> >> > Girish LC
> >> > #09886521716
> >> >
> >>
> >
> >
> >
> > --
> > Regards,
> > Girish LC
> > #09886521716
> >
>



-- 
Regards,
Girish LC
#09886521716

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message