Welcome to the new Gigaspaces XAP forum. To recover your account, please follow these instructions.

Ask Your Question
0

How Backup space detects that Primary is not available?

Hello,

I was under impression that Backup space understands that the Primary is not available, if it does not receive any signals from Primary for some time. But today we have done a test which makes me wonder...

(1) Start Primary + Backup (2) Use UNIX command "kill -STOP <pid>" to suspend the Primary GSC process.

In the Backup GSC log there is a message "channel lost connection to source", but even after a very long time Backup has not taken over a role of Primary.

(3) Use UNIX command "kill -CONT <pid>" to resume the Primary GSC process.

Primary GSC remains primary. In the Backup GSC there is a message "channel reestablished connection to source".

On the other hand, when I really kill the Primary GSC process with the UNIX command "kill -KILL <pid>", the Backup GSC very quickly takes the role of Primary.

My understanding is that when "kill -STOP <pid>" is executed, the affected process will not get any CPU time from the OS, so it will not be able to send anything over the network to the Backup GSC, or to respond to any requests from Backup GSC. The question is why in this case Backup GSC does not take over?

I have done this test with XAP 12.0.1 on SunOS 5.11 on x86, if this plays any role.

asked 2017-08-04 10:16:04 -0600

Alexey Serdyuk gravatar image
edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

The backup will try to ping the primary after it notices that the channel lost connection to the source. because the primary process stopped the backup doesn't get the expected exception and still hangs out waiting for a response. In our testing we use iptables to simulate a disconnection, maybe that approach will work as you expect

answered 2017-08-07 04:15:45 -0600

livnat gravatar image
edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower

Stats

Asked: 2017-08-04 10:16:04 -0600

Seen: 100 times

Last updated: Aug 07