3.22.1.2 upgrade: h2o.ls() now produces java.lang.ArrayIndexOutOfBoundsException: 71

Description

Reported as a SW issue too, but seems to be more of a core h2o problem:

https://github.com/h2oai/sparkling-water/issues/1056

After SW was upgraded to 2.3.22 (and a matching h2o backend cluster version),
h2o.ls() started producing following exception on a freshly created session / freshly created
h2o backend cluster

SW (PySpark) connects to a h2o backend cluster.

Activity

Show:
Ruslan Dautkhanov
February 18, 2019, 5:08 PM

Here's code that reproduces this issue for the upgraded h2o/ SW

I already sent inforamtion how we start h2o backend cluster in email last week.

Let me know if you have further questions

Jakub Hava
February 12, 2019, 10:46 AM

At this stage, it would be great if you could provide us with the information I asked a bit above.

Thank you,
Kuba

Ruslan Dautkhanov
February 11, 2019, 11:49 PM

Sent more information in email earlier today

Let me know if you'd like to jump to a webex if you still can't reproduce it.

Thank you

Jakub Hava
February 11, 2019, 2:23 PM
Edited

Tested on manually started External H2O cluster (1 node) and connected to it from the pysparkling of the correct version. Uploaded a frame and run h2o.ls().

Also killed the client and re-connected with the new client and run h2o.ls(), the output as expected

Jakub Hava
February 11, 2019, 2:05 PM
Edited

I have been testing this today and couldn't reproduce. can you reproduce? If yes, code snippet with reproducible code and shell commands used to start H2O & Sparkling Water would help.

At this point, I believe that our h2o.ls() function works properly but something happened to the h2o cluster

Thank you,
Kuba

Assignee

Unassigned

Fix versions

None

Reporter

Ruslan Dautkhanov

Support ticket URL

None

Labels

None

Affected Spark version

None

Customer Request Type

None

Task progress

None

ReleaseNotesHidden

None

CustomerVisible

No

Priority

Critical