Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
I have a Sense site with two node. Central Scheduler type "Master+ Slave" Second Scheduler "Slave".
If I set Central Scheduler type as "Master" task are not executed.
In system_scheduler.txt log i find it message "No Slaves found to execute Task"
full line
1649 | 20141119T150159.568+0200 | ERROR | Fly | System.Scheduler.Scheduler.Master.Execution.ExecutionManager | 16284 | d726c783-089a-44db-b7a4-7ab159a28d1a | RBCGRP\qvsuser | No Slaves found to execute Task | big data | 8537ccd4-ef4e-4fa3-8be9-0af2ca0e4b4b | RBCGRP\korsikov | 00000000-0000-0000-0000-000000000000 | big data | a34b54b3-422c-43a6-8f18-588f7538c61e | d726c783-089a-44db-b7a4-7ab159a28d1a |
Require additional configuration or a bug server?
Checked. The new version of the server is working correctly. In the default configuration fly (central ) master+slave yukon (second) Slave
start task
Attempting to start task | Reload task of big data |
Sending task Reload task of big data to slavenode fly.rbcgrp.com (Slavenode id: aab50c39-71d3-45fe-9c21-fc9c20d0fb2d)
Reload complete | Reload task of big data |
Engine connection released. 0 of 4 used |
Change config Fly only master yukon only Slave and start task
Attempting to start task | Reload task of big data |
Sending task Reload task of big data to slavenode yukon.rbcgrp.com (Slavenode id: 410304ca-4f37-4db2-b411-faf3bcd8e123) |
Task completed successfully |
Hi Alex,
Thanks for the quick respond.
When I change the scheduler on the master node to Master only - it still works, but as soon as I restart the scheduler service (or the node), stop working. I can see even in QMC the state of the central scheduler changes to stopped.
I don't know if you experience the same.
Thanks,
G
Try to reproduce your situation on my sense cluster. the results are reported immediately
I don't know why is that, but when I switch to Master only it stays in running state. After I reboot the server or restart the service(scheduler), it will change to stopped (maybe it would do it by itself just needed to wait more). When I switch it back to Master and Slave it shows the result in 2 second and turns into running state again.
I will try this on different servers too and see what happens.
my server behaves exactly as you describe. When the state of the scheduler on the central node "stopped" tasks normally performed for the "second" node
Hi,
Just to let you know.
Just testing Sense v 2.0 and this function is working
I could manage to switch te master to master only and the slave to slave only and the scheduled task all reloded on slave node.
This is excellent news!
I'm do not upgrade my test QS cluster to 2.0. Wait for release. Now I am testing 2.0 in a single node.
when the check, be sure to report the results
Hello guys!
Ive just created a 2-node site as a test and im now trying to get my secound node to run the reload task.
If i understand correctly, the "master" is basicly the node that tells the others what to do. The actual reload is always performed by the slave.
As standatd the central noede is both matsre and slave. I installed a second node with just a schedular and added it to the QMC. When starting a task, it gets reloaded on the central node. So i figured i should set the central node to master only. That would make the slave perform the reload. But when i do the task fails with the message: No slave-nodes found to execute Task.
All services are running on both servers and i have tried to restart the shedulers on both servers. Im running QlikSense 2 FR2 (2.2.4.0).
Theese two machines are not connected in an AD and therefor both of them use the local administrator account to run the services. Or do the nodes need to be in the same domain to get this to work?
Any help would be appritiated.
what application do you want to reload? note - monitor application must reloaded only on Central node
Haha it might just be that simple. Since it was a test enviroment, i didnt have any apps so i used the buildin ones
But that should be able to reload on another machine anyways?
EDIT: No, thats not it. I created an app and tried to schedule it. Same issue as befor. ny idees?