Add more logging to discover intermittent RSparkling Issue in jenkins tests

[exalate-issue-sync]

Probably just need higher some specific SparklyR timeouts

{code:java}
10:59:59 > library(testthat)
10:59:59 > library(dplyr)
10:59:59
10:59:59 Attaching package: ‘dplyr’
10:59:59
10:59:59 The following object is masked from ‘package:testthat’:
10:59:59
10:59:59 matches
10:59:59
10:59:59 The following objects are masked from ‘package:stats’:
10:59:59
10:59:59 filter, lag
10:59:59
10:59:59 The following objects are masked from ‘package:base’:
10:59:59 > library(sparklyr)
10:59:59
10:59:59 intersect, setdiff, setequal, union
10:59:59
10:59:59 > library(h2o)
10:59:59
10:59:59 ----------------------------------------------------------------------
10:59:59
10:59:59 Your next step is to start H2O:
10:59:59 > h2o.init()
10:59:59
10:59:59 For H2O package documentation, ask for help:
10:59:59 > ??h2o
10:59:59
10:59:59 After starting H2O, you can use the Web UI at http://localhost:54321
10:59:59 For more information visit http://docs.h2o.ai
10:59:59
10:59:59 ----------------------------------------------------------------------
10:59:59
10:59:59
10:59:59 Attaching package: ‘h2o’
10:59:59
10:59:59 The following objects are masked from ‘package:stats’:
10:59:59
10:59:59 cor, sd, var
10:59:59
10:59:59 The following objects are masked from ‘package:base’:
10:59:59 > library(rsparkling)
10:59:59
10:59:59 %%, %in%, &&, apply, as.factor, as.numeric, colnames, colnames<-,
10:59:59 ifelse, is.character, is.factor, is.numeric, log, log10, log1p,
10:59:59 log2, round, signif, trunc, ||
10:59:59
10:59:59 >
10:59:59 > if (identical(Sys.getenv("NOT_CRAN"), "true")) { # testthat::skip_on_cran
10:59:59 + # Set Sparkling Water Water Jar location and version for tests
10:59:59 + options(rsparkling.sparklingwater.location = Sys.getenv("sparkling.assembly.jar"))
10:59:59 + options(rsparkling.sparklingwater.version = Sys.getenv("sparkling.water.version"))
10:59:59 + # Set Spark version for tests
10:59:59 + options(rsparkling.spark.version = Sys.getenv("spark.version"))
10:59:59 + test_check("rsparkling")
10:59:59 + } else {
10:59:59 + cat("Skipping Tests\n")
10:59:59 + }
11:00:01 ── 1. Error: Test transformation from h2o frame to data frame (@test_transforms.
11:00:01 Failed while connecting to sparklyr to port (8880) for sessionid (67331): ignoring SIGPIPE signal
11:00:01 Path: /home2/jenkins/slave_dir_from_mr-0xc1/workspace/ne_internal_nightly_rel-2.1-AJ6HUL6TB33AKFE3MTI673WEZHGFPSY4PRUARHOD6GTP4HPUGLIA/spark/bin/spark-submit
11:00:01 Parameters: --class, sparklyr.Shell, --jars, '/home2/jenkins/slave_dir_from_mr-0xc1/workspace/ne_internal_nightly_rel-2.1-AJ6HUL6TB33AKFE3MTI673WEZHGFPSY4PRUARHOD6GTP4HPUGLIA/assembly/build/libs/sparkling-water-assembly_2.11-2.1.49-SNAPSHOT-53-all.jar', '/usr/local/lib/R/site-library/sparklyr/java/sparklyr-2.1-2.11.jar', 8880, 67331
11:00:01 Log: /tmp/RtmpIKlxrC/file7d37fe46ea2_spark.log
11:00:01
11:00:01
11:00:01 ---- Output Log ----
11:00:01 Warning: Master yarn-client is deprecated since 2.0. Please use master "yarn" with specified deploy mode instead.
11:00:01 19/02/09 10:00:00 INFO sparklyr: Session (67331) is starting under 127.0.0.1 port 8880
11:00:01 19/02/09 10:00:00 INFO sparklyr: Session (67331) found port 8880 is available
11:00:01 19/02/09 10:00:00 INFO sparklyr: Gateway (67331) is waiting for sparklyr client to connect to port 8880
11:00:01
11:00:01 ---- Error Log ----
11:00:01 1: spark_connect(master = "local[]", config = config) at testthat/test_transforms.R:18
11:00:01 2: shell_connection(master = master, spark_home = spark_home, app_name = app_name, version = version,
11:00:01 hadoop_version = hadoop_version, shell_args = shell_args, config = config, service = spark_config_value(config,
11:00:01 "sparklyr.gateway.service", FALSE), remote = spark_config_value(config, "sparklyr.gateway.remote",
11:00:01 spark_master_is_yarn_cluster(master, config)), extensions = extensions, batch = NULL)
11:00:01 3: start_shell(master = master, spark_home = spark_home, spark_version = version, app_name = app_name,
11:00:01 config = config, jars = spark_config_value(config, c("sparklyr.connect.jars",
11:00:01 "sparklyr.jars.default"), list()), packages = spark_config_value(config,
11:00:01 c("sparklyr.connect.packages", "sparklyr.defaultPackages")), extensions = extensions,
11:00:01 environment = environment, shell_args = shell_args, service = service, remote = remote,
11:00:01 batch = batch)
11:00:01 4: tryCatch({
11:00:01 gatewayInfo <- spark_connect_gateway(gatewayAddress, gatewayPort, sessionId,
11:00:01 config = config, isStarting = TRUE)
11:00:01 }, error = function(e) {
11:00:01 abort_shell(paste("Failed while connecting to sparklyr to port (", gatewayPort,
11:00:01 if (spark_master_is_yarn_cluster(master, config)) {
11:00:01 paste0(") and address (", gatewayAddress)
11:00:01 }
11:00:01 else {
11:00:01 ""
11:00:01 }, ") for sessionid (", sessionId, "): ", e$message, sep = ""), spark_submit_path,
11:00:01 shell_args, output_file, error_file)
11:00:01 })
11:00:01 5: tryCatchList(expr, classes, parentenv, handlers)
11:00:01 6: tryCatchOne(expr, names, parentenv, handlers[[1L]])
11:00:01 7: value[3L]
11:00:01 8: abort_shell(paste("Failed while connecting to sparklyr to port (", gatewayPort, if (spark_master_is_yarn_cluster(master,
11:00:01 config)) {
11:00:01 paste0(") and address (", gatewayAddress)
11:00:01 } else {
11:00:01 ""
11:00:01 }, ") for sessionid (", sessionId, "): ", e$message, sep = ""), spark_submit_path,
11:00:01 shell_args, output_file, error_file)
11:00:01 9: withr::with_options(list(warning.length = 8000), {
11:00:01 maxRows <- 100
11:00:01 Sys.sleep(1)
11:00:01 logLines <- if (!is.null(output_file) && file.exists(output_file)) {
11:00:01 outputContent <- readLines(output_file)
11:00:01 paste(tail(outputContent, n = maxRows), collapse = "\n")
11:00:01 }
11:00:01 else ""
11:00:01 errorLines <- if (!is.null(error_file) && file.exists(error_file))
11:00:01 paste(tail(readLines(error_file), n = maxRows), collapse = "\n")
11:00:01 else ""
11:00:01 stop(paste(message, "\n", if (!is.null(spark_submit_path))
11:00:01 paste(" Path: ", spark_submit_path, "\n", sep = "")
11:00:01 else "", if (!is.null(shell_args))
11:00:01 paste(" Parameters: ", paste(shell_args, collapse = ", "), "\n", sep = "")
11:00:01 else "", if (!is.null(output_file))
11:00:01 paste(" Log: ", output_file, "\n", sep = "")
11:00:01 else "", "\n\n", if (!is.null(output_file))
11:00:01 "---- Output Log ----\n"
11:00:01 else "", logLines, "\n\n", if (!is.null(error_file))
11:00:01 "---- Error Log ----\n"
11:00:01 else "", errorLines, sep = ""))
11:00:01 })
11:00:01 10: force(code)
11:00:01 11: stop(paste(message, "\n", if (!is.null(spark_submit_path)) paste(" Path: ", spark_submit_path,
11:00:01 "\n", sep = "") else "", if (!is.null(shell_args)) paste(" Parameters: ",
11:00:01 paste(shell_args, collapse = ", "), "\n", sep = "") else "", if (!is.null(output_file)) paste(" Log: ",
11:00:01 output_file, "\n", sep = "") else "", "\n\n", if (!is.null(output_file)) "---- Output Log ----\n" else "",
11:00:01 logLines, "\n\n", if (!is.null(error_file)) "---- Error Log ----\n" else "",
11:00:01 errorLines, sep = ""))
11:00:01
11:00:48 ══ testthat results ═══════════════════════════════════════════════════════════
11:00:48 OK: 20 SKIPPED: 0 FAILED: 1
11:00:48 1. Error: Test transformation from h2o frame to data frame (@test_transforms.R#18)
11:00:48
11:00:48 Error: testthat unit tests failed
11:00:48 In addition: Warning message:
11:00:48 In .cacheGenericTable(name, def, .genericTable) :
11:00:48 closing unused connection 3 (->localhost:8880)
11:00:48 Execution halted
11:00:48
11:00:48 > Task :sparkling-water-r:test FAILED
11:00:48
11:00:48 FAILURE: Build failed with an exception.
{code}

[DinukaH2O]

JIRA Issue Migration Info

Jira Issue: SW-1140
Assignee: Jakub Hava
Reporter: Jakub Hava
State: Resolved
Fix Version: 2.1.56
Attachments: N/A
Development PRs: Available

Linked PRs from JIRA

#1277

[hasithjp]

JIRA Issue Migration Info Cont'd

Jira Issue Created Date: 2019-02-12T23:39:33.187-0800