Re: Restarting the Main node when it goes down

[nithin91]

Thanks this information is helpful



--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Re: Restarting the Main node when it goes down

[akorensh]

Ignite will try to connect to each address specified in the addresses section
and establish a connection.
In the case of a multicast ip finder, it will use a multicast group to scan
for nodes.

In general one node needs to discover another and establish a socket
connection on a specified address/port
in order for communication to work.

In your situation where one node has an addresses field and the other one
doesn't, the connection could still be established provided that the
multicast finder can locate the server based on a multicast group.

It is best to put all relevant ips into the addresses section or create a
shared multicast group.

more information here: https://apacheignite.readme.io/docs/tcpip-discovery





--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Re: Restarting the Main node when it goes down

[nithin91]

Also what should the address property have in bean file corresponding to
Client Node.Should it have the list
of all server nodes IP address?Also Can you please explain what happens if
list of all server nodes IP address is specified.



--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Re: Restarting the Main node when it goes down

[nithin91]

Actually in the node A , bean file no IP address is specified for the
property "addresses" corresponding to
class
"org.apache.ignite.spi.discovery.tcp.ipfinder.multicast.TcpDiscoveryMulticastIpFinder"
but in the node B
IP address of A is specified.

Is it because Node A bean file does not have any IP address under address
field , so it is not able to recognize node B even though node B bean file
has IP address corresponding to node A.

Is it required to specify the property "addresses" corresponding to
class "org.apache.ignite.spi.discovery.tcp.ipfinder.multicast" for the main
node .(i.e Node initially started when the launching the ignite cluster)

Also if i have 3 Nodes, what is the value that property "addresses"
corresponding to
class
"org.apache.ignite.spi.discovery.tcp.ipfinder.multicast.TcpDiscoveryMulticastIpFinder"
 
  should have in each of these bean files.Since this property accepts a
list, so do we need to specify all 3 nodes  ip address in each bean file
that is deployed on each node.





--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Re: Restarting the Main node when it goes down

[akorensh]

Hi,
   A baseline topology is set when you activate a cluster w/a set of nodes
the first time around.
  You can subsequently change baseline  the topology using the public API or
control.sh command.

  Once all the baseline topology nodes are up the cluster automatically
activates.

  use control --baseline to see which nodes are in your baseline

  when the cluster restarts  look for:

   Topology snapshot [ver=2, locNode=..., servers=2, clients=0,
state=INACTIVE, CPUs=.., offheap=.., heap=...]
 ^-- Baseline [id=0, size=2, online=2, offline=0]
   ^-- All baseline nodes are online, will start auto-activation

  See: https://apacheignite.readme.io/docs/baseline-topology


   Send the logs of your two nodes and I'll take a look.
Thanks, Alex



--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/

Restarting the Main node when it goes down

[nithin91]

Hi 

I have two nodes Node A(Main Node) and Node B with persistence enabled.When
node B  is restarted whenever it  goes down, it is able to recognize node A
and is joining the Baseline Topology but when Node A alone goes down and
once Node A is restarted, it is not able to recognize Node B and it is not
included in the base line topology.This issue is fixed, only if i stop Node
B and again restarts it.

Information of the baseline Topology that gets displayed on console,
whenever i connect to the cluster as client.

*-- Baseline [id=0, size=2, online=1, offline=1]*

Is there a way to fix this issue with just restarting Node A instead of
stopping and starting the node B.

Is this happening because of enabling Persistence.Also can anyone please let
me know what needs to be done, if the same thing happens without Persistence
being enabled.




--
Sent from: http://apache-ignite-users.70518.x6.nabble.com/