simplify src data env var as we dont use remote data like in db-bench…

…mark v1.0.0

_launcher/solution.R

@@ -117,16 +117,6 @@ file.ext = function(x) {
   if (is.null(ans)) stop(sprintf("solution %s does not have file extension defined in file.ext helper function", x))
   ans
 }
-# data_name env var for each task
-task.env = function(x) {
-  ans = switch(
-    x,
-    "groupby"="SRC_GRP_LOCAL",
-    "join"="SRC_JN_LOCAL"
-  )
-  if (is.null(ans)) stop(sprintf("task %s does not have data name environment variable defined in task.env helper function", x))
-  ans
-}
 # dynamic LHS in: Sys.setenv(var = value)
 setenv = function(var, value, quiet=TRUE) {
   stopifnot(is.character(var), !is.na(var), length(value)==1L, is.atomic(value))
@@ -145,9 +135,6 @@ data.desc = function(task, nrow, k, na, sort) {
   }
   sprintf("%s_%s_%s_%s_%s", prefix, nrow, k, na, sort)
 }
-data_name_exception = function(solution, task, d) {
-  d
-}
 # no dots solution name used in paths
 solution.path = function(x) {
   gsub(".", "", x, fixed=TRUE)
@@ -157,23 +144,21 @@ solution.path = function(x) {
 
 s = args[["solution"]]
 t = args[["task"]]
-data_name_env = task.env(t)
 d = data.desc(t, args[["nrow"]], args[["k"]], args[["na"]], args[["sort"]])
-d = data_name_exception(solution=s, task=t, d=d) # this is already handled in launch.R but here we handle ad-hoc single solution cmd runs
 
 Sys.setenv("CSV_TIME_FILE"=args[["out"]])
-setenv(data_name_env, d)
+setenv("SRC_DATANAME", d)
 
 ns = solution.path(s)
 ext = file.ext(s)
 localcmd = if (s %in% c("clickhouse","h2o")) { # custom launcher bash script, for clickhouse and h2o
-  sprintf("exec.sh %s %s", t, d)
+  sprintf("exec.sh %s", t)
 } else sprintf("%s-%s.%s", t, ns, ext)
 cmd = sprintf("./%s/%s", ns, localcmd)
 
 ret = system(cmd, ignore.stdout=as.logical(args[["quiet"]]))
 
-Sys.unsetenv(data_name_env)
+Sys.unsetenv("SRC_DATANAME")
 Sys.unsetenv("CSV_TIME_FILE")
 
 if (stdout && file.size(args[["out"]])) {

clickhouse/exec.sh

@@ -1,9 +1,9 @@
 #!/bin/bash
 set -e
 
-if [ "$#" -ne 2 ]; then
-    echo 'usage: ./clickhouse/exec.sh groupby G1_1e7_1e2_0_0';
-    exit 1
+if [ "$#" -ne 1 ]; then
+  echo 'usage: ./clickhouse/exec.sh groupby';
+  exit 1
 fi;
 
 source ./clickhouse/ch.sh
@@ -19,56 +19,56 @@ ch_active || exit 1
 
 # tune CH settings and load data
 CH_MEM=107374182400 # 100GB ## old value 128849018880 # 120GB ## now set to 96GB after cache=1 to in-memory temp tables because there was not enough mem for R to parse timings
-clickhouse-client --query "DROP TABLE IF EXISTS ans"
-echo "# clickhouse/exec.sh: creating tables and loading data"
+clickhouse-client --query 'DROP TABLE IF EXISTS ans'
+echo '# clickhouse/exec.sh: creating tables and loading data'
 if [ $1 == 'groupby' ]; then
   CH_EXT_GRP_BY=53687091200 # twice less than CH_MEM #96
   CH_EXT_SORT=53687091200
-  clickhouse-client --query "DROP TABLE IF EXISTS $2"
-  clickhouse-client --query "CREATE TABLE $2 (id1 String, id2 String, id3 String, id4 Int32, id5 Int32, id6 Int32, v1 Int32, v2 Int32, v3 Float64) ENGINE = MergeTree() ORDER BY tuple();"
-  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $2 SELECT id1, id2, id3, id4, id5, id6, v1, v2, v3 FROM file('data/$2.csv', 'CSVWithNames', 'id1 String, id2 String, id3 String, id4 Int32, id5 Int32, id6 Int32, v1 Int32, v2 Int32, v3 Float64')"
+  clickhouse-client --query "DROP TABLE IF EXISTS $SRC_DATANAME"
+  clickhouse-client --query "CREATE TABLE $SRC_DATANAME (id1 String, id2 String, id3 String, id4 Int32, id5 Int32, id6 Int32, v1 Int32, v2 Int32, v3 Float64) ENGINE = MergeTree() ORDER BY tuple();"
+  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $SRC_DATANAME SELECT id1, id2, id3, id4, id5, id6, v1, v2, v3 FROM file('data/$SRC_DATANAME.csv', 'CSVWithNames', 'id1 String, id2 String, id3 String, id4 Int32, id5 Int32, id6 Int32, v1 Int32, v2 Int32, v3 Float64')"
   # confirm all data loaded yandex/ClickHouse#4463
-  echo -e "clickhouse-client --query 'SELECT count(*) FROM $2'\n$2" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(strsplit(stdin[2L], "_", fixed=TRUE)[[1L]][2L])) stop("incomplete data load for ", stdin[2L],", loaded ", loaded, " rows only")'
+  echo -e "clickhouse-client --query 'SELECT count(*) FROM $SRC_DATANAME'\n$(echo $SRC_DATANAME | cut -d'_' -f2)" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(stdin[2L])) stop("incomplete data load, expected: ", stdin[2L],", loaded: ", loaded)'
 elif [ $1 == 'join' ]; then
   # lhs
-  clickhouse-client --query "DROP TABLE IF EXISTS $2"
-  clickhouse-client --query "CREATE TABLE $2 (id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v1 Float64) ENGINE = MergeTree() ORDER BY tuple();"
-  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $2 SELECT id1, id2, id3, id4, id5, id6, v1 FROM file('data/$2.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v1 Float64')"
-  echo -e "clickhouse-client --query 'SELECT count(*) FROM $2'\n$2" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(strsplit(stdin[2L], "_", fixed=TRUE)[[1L]][2L])) stop("incomplete data load for ", stdin[2L],", loaded ", loaded, " rows only")'
-  rhs=$(join_to_tbls $2)
-  rhs1=$(echo $rhs | cut -d ' ' -f1)
-  clickhouse-client --query "DROP TABLE IF EXISTS $rhs1"
-  clickhouse-client --query "CREATE TABLE $rhs1 (id1 Int32, id4 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
-  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $rhs1 SELECT id1, id4, v2 FROM file('data/$rhs1.csv', 'CSVWithNames', 'id1 Int32, id4 String, v2 Float64')"
-  echo -e "clickhouse-client --query 'SELECT count(*) FROM $rhs1'\n$rhs1" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(strsplit(stdin[2L], "_", fixed=TRUE)[[1L]][3L])) stop("incomplete data load for ", stdin[2L],", loaded ", loaded, " rows only")'
-  rhs2=$(echo $rhs | cut -d ' ' -f2)
-  clickhouse-client --query "DROP TABLE IF EXISTS $rhs2"
-  clickhouse-client --query "CREATE TABLE $rhs2 (id1 Int32, id2 Int32, id4 String, id5 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
-  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $rhs2 SELECT id1, id2, id4, id5, v2 FROM file('data/$rhs2.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id4 String, id5 String, v2 Float64')"
-  echo -e "clickhouse-client --query 'SELECT count(*) FROM $rhs2'\n$rhs2" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(strsplit(stdin[2L], "_", fixed=TRUE)[[1L]][3L])) stop("incomplete data load for ", stdin[2L],", loaded ", loaded, " rows only")'
-  rhs3=$(echo $rhs | cut -d ' ' -f3)
-  clickhouse-client --query "DROP TABLE IF EXISTS $rhs3"
-  clickhouse-client --query "CREATE TABLE $rhs3 (id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
-  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $rhs3 SELECT id1, id2, id3, id4, id5, id6, v2 FROM file('data/$rhs3.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v2 Float64')"
-  echo -e "clickhouse-client --query 'SELECT count(*) FROM $rhs3'\n$rhs3" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(strsplit(stdin[2L], "_", fixed=TRUE)[[1L]][3L])) stop("incomplete data load for ", stdin[2L],", loaded ", loaded, " rows only")'
+  clickhouse-client --query "DROP TABLE IF EXISTS $SRC_DATANAME"
+  clickhouse-client --query "CREATE TABLE $SRC_DATANAME (id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v1 Float64) ENGINE = MergeTree() ORDER BY tuple();"
+  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $SRC_DATANAME SELECT id1, id2, id3, id4, id5, id6, v1 FROM file('data/$SRC_DATANAME.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v1 Float64')"
+  echo -e "clickhouse-client --query 'SELECT count(*) FROM $SRC_DATANAME'\n$(echo $SRC_DATANAME | cut -d'_' -f2)" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(stdin[2L])) stop("incomplete data load, expected: ", stdin[2L],", loaded: ", loaded)'
+  RHS=$(join_to_tbls $SRC_DATANAME)
+  RHS1=$(echo $RHS | cut -d' ' -f1)
+  clickhouse-client --query "DROP TABLE IF EXISTS $RHS1"
+  clickhouse-client --query "CREATE TABLE $RHS1 (id1 Int32, id4 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
+  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $RHS1 SELECT id1, id4, v2 FROM file('data/$RHS1.csv', 'CSVWithNames', 'id1 Int32, id4 String, v2 Float64')"
+  echo -e "clickhouse-client --query 'SELECT count(*) FROM $RHS1'\n$(echo $RHS1 | cut -d'_' -f3)" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(stdin[2L])) stop("incomplete data load, expected: ", stdin[2L],", loaded: ", loaded)'
+  RHS2=$(echo $RHS | cut -d' ' -f2)
+  clickhouse-client --query "DROP TABLE IF EXISTS $RHS2"
+  clickhouse-client --query "CREATE TABLE $RHS2 (id1 Int32, id2 Int32, id4 String, id5 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
+  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $RHS2 SELECT id1, id2, id4, id5, v2 FROM file('data/$RHS2.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id4 String, id5 String, v2 Float64')"
+  echo -e "clickhouse-client --query 'SELECT count(*) FROM $RHS2'\n$(echo $RHS2 | cut -d'_' -f3)" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(stdin[2L])) stop("incomplete data load, expected: ", stdin[2L],", loaded: ", loaded)'
+  RHS3=$(echo $RHS | cut -d' ' -f3)
+  clickhouse-client --query "DROP TABLE IF EXISTS $RHS3"
+  clickhouse-client --query "CREATE TABLE $RHS3 (id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v2 Float64) ENGINE = MergeTree() ORDER BY tuple();"
+  clickhouse-client --max_memory_usage $CH_MEM --max_insert_threads 1 --query "INSERT INTO $RHS3 SELECT id1, id2, id3, id4, id5, id6, v2 FROM file('data/$RHS3.csv', 'CSVWithNames', 'id1 Int32, id2 Int32, id3 Int32, id4 String, id5 String, id6 String, v2 Float64')"
+  echo -e "clickhouse-client --query 'SELECT count(*) FROM $RHS3'\n$(echo $RHS3 | cut -d'_' -f3)" | Rscript -e 'stdin=readLines(file("stdin")); if ((loaded<-as.numeric(system(stdin[1L], intern=TRUE)))!=as.numeric(stdin[2L])) stop("incomplete data load, expected: ", stdin[2L],", loaded: ", loaded)'
 else
   echo "clickhouse task $1 not implemented" >&2 && exit 1
 fi
 
 # cleanup timings from last run if they have not been cleaned up after parsing
 mkdir -p clickhouse/log
-rm -f clickhouse/log/$1_$2_q*.csv
+rm -f clickhouse/log/$1_${SRC_DATANAME}_q*.csv
 
 # execute sql script on clickhouse
-clickhouse-client --query "TRUNCATE TABLE system.query_log"
-echo "# clickhouse/exec.sh: data loaded, logs truncated, preparing $1-$2 benchmark sql script and sending it clickhouse"
+clickhouse-client --query 'TRUNCATE TABLE system.query_log'
+echo "# clickhouse/exec.sh: data loaded, logs truncated, preparing $1-$SRC_DATANAME benchmark sql script and sending it clickhouse"
 if [ $1 == 'groupby' ]; then
   # for each data_name produce sql script
-  sed "s/DATA_NAME/$2/g" < "clickhouse/$1-clickhouse.sql.in" > "clickhouse/$1-clickhouse.sql"
-  cat "clickhouse/$1-clickhouse.sql" | clickhouse-client -mn --max_memory_usage $CH_MEM --max_bytes_before_external_group_by $CH_EXT_GRP_BY --max_bytes_before_external_sort $CH_EXT_SORT --receive_timeout 10800 --format Pretty --output_format_pretty_max_rows 1 && echo '# clickhouse/exec.sh: benchmark sql script finished' || echo "# clickhouse/exec.sh: benchmark sql script for $2 terminated with error"
+  sed "s/DATA_NAME/$SRC_DATANAME/g" < "clickhouse/$1-clickhouse.sql.in" > "clickhouse/$1-clickhouse.sql"
+  cat "clickhouse/$1-clickhouse.sql" | clickhouse-client -mn --max_memory_usage $CH_MEM --max_bytes_before_external_group_by $CH_EXT_GRP_BY --max_bytes_before_external_sort $CH_EXT_SORT --receive_timeout 10800 --format Pretty --output_format_pretty_max_rows 1 && echo '# clickhouse/exec.sh: benchmark sql script finished' || echo "# clickhouse/exec.sh: benchmark sql script for $SRC_DATANAME terminated with error"
 elif [ $1 == 'join' ]; then
-  sed "s/DATA_NAME/$2/g; s/RHS_SMALL/$rhs1/g; s/RHS_MEDIUM/$rhs2/g; s/RHS_BIG/$rhs3/g" < "clickhouse/join-clickhouse.sql.in" > "clickhouse/join-clickhouse.sql"
-  cat "clickhouse/$1-clickhouse.sql" | clickhouse-client -mn --max_memory_usage $CH_MEM --receive_timeout 10800 --format Pretty --output_format_pretty_max_rows 1 && echo '# clickhouse/exec.sh: benchmark sql script finished' || echo "# clickhouse/exec.sh: benchmark sql script for $2 terminated with error"
+  sed "s/DATA_NAME/$SRC_DATANAME/g; s/RHS_SMALL/$RHS1/g; s/RHS_MEDIUM/$RHS2/g; s/RHS_BIG/$RHS3/g" < "clickhouse/join-clickhouse.sql.in" > "clickhouse/join-clickhouse.sql"
+  cat "clickhouse/$1-clickhouse.sql" | clickhouse-client -mn --max_memory_usage $CH_MEM --receive_timeout 10800 --format Pretty --output_format_pretty_max_rows 1 && echo '# clickhouse/exec.sh: benchmark sql script finished' || echo "# clickhouse/exec.sh: benchmark sql script for $SRC_DATANAME terminated with error"
 else
   echo "clickhouse task $1 benchmark script launching not defined" >&2 && exit 1
 fi
@@ -77,7 +77,7 @@ fi
 sleep 90
 
 # cleanup data
-ch_active && echo '# clickhouse/exec.sh: finishing, cleaning up' && clickhouse-client --query "DROP TABLE IF EXISTS ans" && clickhouse-client --query "DROP TABLE IF EXISTS $2" || echo '# clickhouse/exec.sh: finishing, clickhouse server down, possibly crashed, could not clean up'
+ch_active && echo '# clickhouse/exec.sh: finishing, cleaning up' && clickhouse-client --query "DROP TABLE IF EXISTS ans" && clickhouse-client --query "DROP TABLE IF EXISTS $SRC_DATANAME" || echo '# clickhouse/exec.sh: finishing, clickhouse server down, possibly crashed, could not clean up'
 
 # stop server
 ch_stop && echo '# clickhouse/exec.sh: stopping server finished' || echo '# clickhouse/exec.sh: stopping server failed'
@@ -86,4 +86,4 @@ ch_stop && echo '# clickhouse/exec.sh: stopping server finished' || echo '# clic
 sleep 30
 
 # parse timings from clickhouse/log/[task]_[data_name]_q[i]_r[j].csv
-Rscript clickhouse/clickhouse-parse-log.R "$1" "$2"
+Rscript clickhouse/clickhouse-parse-log.R $1 $SRC_DATANAME

cudf/groupby-cudf.py

@@ -16,7 +16,7 @@
 fun = ".groupby"
 cache = "TRUE"
 
-data_name = os.environ['SRC_GRP_LOCAL']
+data_name = os.environ['SRC_DATANAME']
 src_grp = os.path.join("data", data_name+".csv")
 print("loading dataset %s" % data_name, flush=True)
 

cudf/join-cudf.py

@@ -16,7 +16,7 @@
 fun = ".merge"
 cache = "TRUE"
 
-data_name = os.environ['SRC_JN_LOCAL']
+data_name = os.environ['SRC_DATANAME']
 src_jn_x = os.path.join("data", data_name+".csv")
 y_data_name = join_to_tbls(data_name)
 src_jn_y = [os.path.join("data", y_data_name[0]+".csv"), os.path.join("data", y_data_name[1]+".csv"), os.path.join("data", y_data_name[2]+".csv")]

dask/groupby-dask2.py

@@ -25,7 +25,7 @@
 # since we are running on local cluster of processes, we would prefer to keep the communication between workers to relative minimum, thus it's better to trade some tasks granularity for better processing locality
 dk.config.set({"optimization.fuse.ave-width": 20})
 
-data_name = os.environ['SRC_GRP_LOCAL']
+data_name = os.environ['SRC_DATANAME']
 on_disk = False #data_name.split("_")[1] == "1e9" # on-disk data storage #126
 fext = "parquet" if on_disk else "csv"
 src_grp = os.path.join("data", data_name+"."+fext)

dask/join-dask.py

@@ -18,7 +18,7 @@
 fun = ".merge"
 cache = "TRUE"
 
-data_name = os.environ['SRC_JN_LOCAL']
+data_name = os.environ['SRC_DATANAME']
 on_disk = False #data_name.split("_")[1] == "1e9" # on-disk data storage #126
 fext = "parquet" if on_disk else "csv"
 src_jn_x = os.path.join("data", data_name+"."+fext)

datatable/groupby-datatable.R

@@ -15,7 +15,7 @@ fun = "[.data.table"
 cache = TRUE
 on_disk = FALSE
 
-data_name = Sys.getenv("SRC_GRP_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_grp = file.path("data", paste(data_name, "csv", sep="."))
 #src_grp = file.path("data", paste(data_name, "rds", sep="."))
 cat(sprintf("loading dataset %s\n", data_name))

datatable/join-datatable.R

@@ -14,7 +14,7 @@ fun = "[.data.table"
 cache = TRUE
 on_disk = FALSE
 
-data_name = Sys.getenv("SRC_JN_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_jn_x = file.path("data", paste(data_name, "csv", sep="."))
 y_data_name = join_to_tbls(data_name)
 src_jn_y = setNames(file.path("data", paste(y_data_name, "csv", sep=".")), names(y_data_name))

dplyr/groupby-dplyr.R

@@ -15,7 +15,7 @@ fun = "group_by"
 cache = TRUE
 on_disk = FALSE
 
-data_name = Sys.getenv("SRC_GRP_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_grp = file.path("data", paste(data_name, "csv", sep="."))
 cat(sprintf("loading dataset %s\n", data_name))
 

dplyr/join-dplyr.R

@@ -14,7 +14,7 @@ solution = "dplyr"
 cache = TRUE
 on_disk = FALSE
 
-data_name = Sys.getenv("SRC_JN_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_jn_x = file.path("data", paste(data_name, "csv", sep="."))
 y_data_name = join_to_tbls(data_name)
 src_jn_y = setNames(file.path("data", paste(y_data_name, "csv", sep=".")), names(y_data_name))

h2o/exec.sh

@@ -1,28 +1,27 @@
 #!/bin/bash
 set -e
 
-if [ "$#" -ne 2 ]; then
-    echo "usage: ./h2o/exec.sh groupby G1_1e7_1e2_0_0";
-    exit 1
+if [ "$#" -ne 1 ]; then
+  echo 'usage: ./h2o/exec.sh groupby';
+  exit 1
 fi;
 
 source ./h2o/h2o.sh
 
-h2o_active && echo "h2o instance should not be already running, investigate" >&2
+h2o_active && echo 'h2o instance should not be already running, investigate' >&2
 h2o_active && exit 1
 
 # start h2o
-h2o_start "h2o_$1_$2"
+h2o_start "h2o_$1_""$SRC_DATANAME"
 
 # confirm h2o working
 h2o_active || sleep 30
-h2o_active || echo "h2o instance should be already running, investigate" >&2
+h2o_active || echo 'h2o instance should be already running, investigate' >&2
 h2o_active || exit 1
 
 # execute benchmark script
-./h2o/$1-h2o.R || echo "# h2o/exec.sh: benchmark script for $2 terminated with error" >&2
+./h2o/$1-h2o.R || echo "# h2o/exec.sh: benchmark script for $SRC_DATANAME terminated with error" >&2
 
 # stop h2o instance
-h2o_stop && echo "# h2o/exec.sh: stopping h2o instance finished" || echo "# h2o/exec.sh: stopping h2o instance failed" >&2
+h2o_stop && echo '# h2o/exec.sh: stopping h2o instance finished' || echo '# h2o/exec.sh: stopping h2o instance failed' >&2
 h2o_active || exit 1
-

h2o/groupby-h2o.R

@@ -16,7 +16,7 @@ on_disk = FALSE
 h = h2o.init(startH2O=FALSE, port=55888)
 h2o.no_progress()
 
-data_name = Sys.getenv("SRC_GRP_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_grp = file.path("data", paste(data_name, "csv", sep="."))
 cat(sprintf("loading dataset %s\n", data_name))
 

h2o/join-h2o.R

@@ -16,7 +16,7 @@ on_disk = FALSE
 h = h2o.init(startH2O=FALSE, port=55888)
 h2o.no_progress()
 
-data_name = Sys.getenv("SRC_JN_LOCAL")
+data_name = Sys.getenv("SRC_DATANAME")
 src_jn_x = file.path("data", paste(data_name, "csv", sep="."))
 y_data_name = join_to_tbls(data_name)
 src_jn_y = setNames(file.path("data", paste(y_data_name, "csv", sep=".")), names(y_data_name))

juliadf/groupby-juliadf.jl

@@ -21,7 +21,7 @@ fun = "by";
 cache = true;
 on_disk = false;
 
-data_name = ENV["SRC_GRP_LOCAL"];
+data_name = ENV["SRC_DATANAME"];
 src_grp = string("data/", data_name, ".csv");
 println(string("loading dataset ", data_name)); flush(stdout);
 

juliadf/join-juliadf.jl

@@ -20,7 +20,7 @@ fun = "join";
 cache = true;
 on_disk = false;
 
-data_name = ENV["SRC_JN_LOCAL"];
+data_name = ENV["SRC_DATANAME"];
 src_jn_x = string("data/", data_name, ".csv");
 y_data_name = join_to_tbls(data_name);
 src_jn_y = [string("data/", y_data_name[1], ".csv"), string("data/", y_data_name[2], ".csv"), string("data/", y_data_name[3], ".csv")];

modin/groupby-modin.py

@@ -18,7 +18,7 @@
 cache = "TRUE"
 on_disk = "FALSE"
 
-data_name = os.environ['SRC_GRP_LOCAL']
+data_name = os.environ['SRC_DATANAME']
 src_grp = os.path.join("data", data_name+".csv")
 print("loading dataset %s" % data_name, flush=True)