Thread tcp-disco-srvr-nextflow is terminated unexpectedly.

[nephantes]

Bug report

I started a cluster with a master and a worker node. Cluster is terminated unexpectedly.
The image I used derived from your example image.

Expected behavior and actual behavior

It shouldn't be terminated.

Program output

Logs from master node;
https://galaxyweb.umassmed.edu/logs/test6/master/.nextflow.log
Other logs;
https://galaxyweb.umassmed.edu/logs/test6/master/

Logs from worker node;
https://galaxyweb.umassmed.edu/logs/test6/worker/.node-nextflow.log
Other logs;
https://galaxyweb.umassmed.edu/logs/test6/worker/

Steps to reproduce the problem

Create 2 nodes cluster with ami-0a60f8f1463a043b4 on eu-west1b.
When I run the pipeline below with more than 20 fastq pairs. I got this error in 2 nodes cluster.
https://galaxyweb.umassmed.edu/logs/test6/test/

Environment

• Nextflow version: nextflow version 0.32.0.4897
• Java version: openjdk version "1.8.0_181"
• Operating system: Amazon Linux AMI release 2018.03

[pditommaso]

It looks an Ignite bug. https://issues.apache.org/jira/browse/IGNITE-8751

You may want to try to downgrade to NF 0.31.1 that uses Ignite 2.4.

[nephantes]

It looks like previous version is working. I didn't get the same error.

[nephantes]

Hi Paolo,
This issue looks like resolved.
https://issues.apache.org/jira/browse/IGNITE-8751
Which nextflow version, we can start using this ignites resolved version.
Thanks

[pditommaso]

Yes, but it has not yet released. After that it will include in the first version according the new nextflow release schema. Alternatively you can rollback to an older NF release which uses a previous Ignite version.

[pditommaso]

Reverted to Ignite 2.4.

[pditommaso]

Included in version 18.12.0-edge.