diff --git a/convert.py b/convert.py
new file mode 100644
index 0000000000000000000000000000000000000000..f448cf034cc869097e31c1a224bffe13905ad494
--- /dev/null
+++ b/convert.py
@@ -0,0 +1,79 @@
+import numpy
+import pandas
+from datetime import datetime
+import time
+
+
+def datetoint(col):
+	return numpy.array([time.mktime(datetime.strptime(x, '%Y-%m-%d').timetuple()) for x in col], dtype=numpy.int64)
+
+
+cfile = pandas.read_csv('lineitem-1m.tbl', sep='|', header=None, names=['l_orderkey','l_partkey','l_suppkey','l_linenumber','l_quantity','l_extendedprice','l_discount','l_tax','l_returnflag','l_linestatus','l_shipdate','l_commitdate','l_receiptdate','l_shipinstruct','l_shipmode','l_comment'])
+
+cfile.l_orderkey = cfile.l_orderkey.astype(numpy.int32)
+cfile.l_partkey = cfile.l_partkey.astype(numpy.int32)
+cfile.l_suppkey = cfile.l_suppkey.astype(numpy.int16)
+cfile.l_linenumber = cfile.l_linenumber.astype(numpy.int8)
+cfile.l_quantity = cfile.l_quantity.astype(numpy.int8)
+cfile.l_shipdate = datetoint(cfile.l_shipdate)
+cfile.l_commitdate = datetoint(cfile.l_commitdate)
+cfile.l_receiptdate = datetoint(cfile.l_receiptdate)
+
+
+chunkbytes=1000000
+
+tmp = chunkbytes
+tmp -= 16
+row = 0
+startrow = 0
+
+rowsize = 0
+
+while True:
+	rowsize = 0
+	for i in range(len(cfile.columns)):
+		col = cfile[cfile.columns[i]]
+		if col.dtype == numpy.object:
+			rowsize += len(col[row]) + 1
+		else:
+			rowsize += col.dtype.itemsize
+	if (tmp < rowsize):
+		break
+	tmp -= rowsize
+	row += 1
+
+
+out = open('out.col', 'wb+')
+
+# message length
+numpy.array(chunkbytes - tmp, dtype=numpy.int64).tofile(out)
+# row count
+numpy.array(row, dtype=numpy.int64).tofile(out)
+
+for i in range(len(cfile.columns)):
+	col = cfile[cfile.columns[i]]
+	if col.dtype == numpy.object:
+		for o in range(startrow, (startrow + row)):
+			out.write(col[o])
+			out.write('\0')
+	else:
+		numpy.array(col[startrow:(startrow + row)]).tofile(out)
+
+out.close()
+
+
+
+out = open('out.row', 'wb+')
+
+# header
+numpy.array(chunkbytes - tmp, dtype=numpy.int64).tofile(out)
+for r in range(startrow, startrow+row):
+	for i in range(len(cfile.columns)):
+		col = cfile[cfile.columns[i]]
+		if col.dtype == numpy.object:
+			out.write(col[r])
+			out.write('\0')
+		else:
+			numpy.array(col[r]).tofile(out)
+
+out.close()
diff --git a/pmjc.jar b/pmjc.jar
new file mode 100644
index 0000000000000000000000000000000000000000..4d4bc05878b4bbbf25f78e0874bd4da91c4ac69b
GIT binary patch
literal 2087
zcmZ{l2T+sg7KTF!<c1cCNXWVp>ADbUE=6epObCI5j;OSR5~QObMKN#z5tQB{2na+&
z2bW7xBp{s)1S7Z<MKlN~Z4oR8Y_c==!rmF*nZJDV&v(xMerM)=a2P&*DF6rr0<8J-
zq5%ga1>gr@&FxLp(AE|Rtcf++!rabY9c!^W2mowghli{%YU)f8jM{O=@JN&GNtV_u
zdydTuz+nVIjn{csV4PA=&V~D50~GO}4M3c)>cOEwB%)XN<#eoQmmK8Z6H(yjvA0&v
zkG73Q@@xV?S&5b32_<QL5xxx#N-M9THxk8Wsd<YL^6sOI@t)Gj!v^(ZssU4Nk<)eT
z23;N373O8Rw?MMlvV0C@InmXlWS!yRmmor<7B^-;5rFAYyjd9x8L%wFFr_QzJgP93
z^1dYc@`l_1A)->p!vU<x^l;?~iHW1xHDrpSlkrxuoj_|L-YGb~`_dnSVOz+4lMVoL
zF7Bk0&(6IwBa%@|6-~LD?6j)AXg;oUPgUP*&p5tfcn&&W`oE*)o=e~*gja+gg5ZG&
z4j_3UaGaaE*Cmgzur!A$JD7vaUQzu(BB3-uE}-;Lg~KHY%lmjdsOGSW1Xe{+?ZvT|
z#4wN$Wrl#5t{!JBRNc}4W*_^oZcg}=AuD2!XYVjo{+f?!>1cTG>AOwu+pj-e&3wAM
z^Hw8nBE-!iuC!ZwY_hOe{ox{s=z_6f!jYc3ZPm3jKcvpkqQB74`7o%deA~lzCz*2+
z15g1!*70@Fxlx*2GNe1>ap0fmy2j3B*b<oz^}LF)OM|``@7Y$xhvtPR!Ha|D^f4b(
zXcXZ-$)>M6kd)dQ3cETO*4kFPPOEaQl$Y-vZAa*k9^BUQYkiyRlPl|~sXstLuvw|$
z>rowaQcuX;?2tL*wjPl__95Vfg-ZL@fLqS7&4CqE716x1HzY-vg&1<H@_z2u;MGe5
z<4&z-CJa9Ok*IrQ4_}cy{!6>rJ)JvKuy91ElWXWFtl?og^wXd&hUn|*M%TORlU_w8
zKor)^w9Jl0LhJe$<iCr-OJ{i{2=2^A<riT1>xp6G-E_xOWGVH*1tU-jqQ)JNX9f&-
z(luIuG7LhbI-xs12_s()sIdHka39tRsj`gHpRT}8P#zzl%e~57m(F>Dr)nG>A(JE?
zH<RJQT90yC4M+t_e)vrpFjnH5Q8qP1=-2J}MttgQO6pKFTeshFP4tywx(+RE=$87C
zLxF8&cK0+UN+eTPy{vjwx*KBm^q#Y?={spKAhopS?yxS@^U3+L{LUoJP;K-`G;U4&
zPvMmAnGmL}IZ2gln1Nw{o}hsfC(UCz7&;^D3G;0Yg^SaCUEj1LPIs$?=8fh{cibBJ
z78nfNs3b$DJ(F$m;Aw|Gdncl9malif9UDg@dUfiQ@C+i?c1?9&W^#uFyV&zn#f>XF
zPqHNCw+qMyyTRv*j}DM|MHCYqZC_&1fI|PVBF{PLL)ks<z#Tdm>Ej;J?Ob}|ST%FD
zpx#V!WAg%4s-}FBNP2x+33C&UyCB(K`>(TO53b8@zr{Tjp9(qguB^#rFWqHIuXCy{
z4P@OUoRWVo9?u(4POXt*(Vlt{S-v*XY9r%cTGX!uRyo(^B`BUdo-O$dVeIy(Wifj(
zr;@@vVzqO7t<KQE)Tdv$`IsfO@q}>%#m2u+JX3!O0}K{E(gleI4E*}Vbt&YFkJ*19
zcZw!^7ws|g;uqU;z{KW^;m>_zQ<2Mp{4R!yL|eMmp8R8%7H2{BDZO-X*2n0}F_2DU
zO9l%Ma(7<J+*+96TgY5$q2v1cgyZQ+Ez_DZQOVDw=iAF8QEhffvYn?QPfLkeDjMDX
z`cgu?LsY|L+wXmliHlOoNpt7jmJKSZy~if{XXWn+$ZQ%)+cLtSLYAr>4M84NjJ|OZ
znhEPd7rqa)dLuO-EPH}QP-1kMBRb%ZpY~jhIlC%YzA3ccjI!(0YY>v8Jz&nnuX=RD
zleJoG+I}|CPrz1trg;{YiMsU(n6#TlQNG+A#N98Psi=c=$n7Y++ftr;{mXChAqj*S
z^gT@rf>nQhW!D*a{RnWkR9j}{>*mgq5jV}i0{OR+_yGVBPE`Lm$HCkj|2w*ZJ;Hrs
z?_kB~#*oC3?<wA^MWJ(}5*e1-CZ?@}S0ZeO3In!~Up|9O6{fwFMkrTbg-04LEhLPq
zBB38llgv?FDD%s`<7|bMrZ*);;NOe<TV9Jy)_f+vTUOLp$%KYE)-PO(Ngc4k$wb}g
z9gj1u>V<!-5d59*=FGB=)KHAU#o-}lcrhX?$U5UJ=6w9K<17eE>!!40an?J>KyP*h
zy?j==WALr|r(KzXS<0-NZBCg1Md@f!9oMY?BQokYjuqv<qnn3U3b4<`gFxeQ@t^t;
zC2^SjDqQ#-gaa3T`|{wt{uhM@cX7dSa0py*?8}w&r$5BV!7|+D2kmhuwl7tIgY&~-
TfSfgeW0D`R&3Q@-a#w!=P{e2u

literal 0
HcmV?d00001

diff --git a/vldb-protocols.py b/vldb-protocols.py
index c5c2114e5aca61ae49ead94b4190d9f7929c67ee..6a4a0d0a193efed6805665da55c73aba0e6bcd3d 100644
--- a/vldb-protocols.py
+++ b/vldb-protocols.py
@@ -7,6 +7,7 @@ import json
 import subprocess
 
 systems = [
+	{'name':'hive-default', 'db':'hive'}, 
 	{'name':'netcat-csv-lz4', 'db':'netcat', 'compress': 'lz4'},
 	{'name':'netcat-csv-lz4-heavy', 'db':'netcat', 'compress': 'lz4-heavy'},
 	{'name':'netcat-csv-gzip', 'db':'netcat', 'compress': 'gzip'},
@@ -18,7 +19,7 @@ systems = [
 	{'name':'postgres-default', 'db':'postgres'}, 
 	{'name':'mariadb-default', 'db':'mariadb'}, 
 	{'name':'monetdb-default', 'db':'monetdb'}, 
-	{'name':'hbase-default', 'db':'hbase'}, 
+	{'name':'hbase-default', 'db':'hbase'},
 	{'name':'mongodb-default', 'db':'mongodb'}]
 
 networks = [
@@ -123,28 +124,28 @@ for r in range(nruns):
 				odbccmd = None
 				if system['db'] == 'postgres':
 					querycmd = 'psql %s --host 127.0.0.1 -w -t -A -c "%s" > /dev/null' % ('--set=sslcompression=1 --set=sslmode=require --set=keepalives=0' if 'compress' in system else '', query)
-					jdbcflags = '-u jdbc:postgresql://127.0.0.1/user -n user -d org.postgresql.Driver'
+					jdbcflags = 'org.postgresql.Driver jdbc:postgresql://127.0.0.1/user user user'
 					odbccmd = 'isql PostgreSQL -d, < query > /dev/null'
 				elif system['db'] == 'mariadb':
 					querycmd = 'mysql %s --host=127.0.0.1 user --skip-column-names --batch -e "%s" > /dev/null'  % ('--compress' if 'compress' in system else '', query)
-					jdbcflags = '-u jdbc:mysql://127.0.0.1/user -n user -d org.mariadb.jdbc.Driver'
+					jdbcflags = 'org.mariadb.jdbc.Driver jdbc:mysql://127.0.0.1/user user null'
 					odbccmd = 'isql MySQL -d, < query > /dev/null'
 				elif system['db'] == 'monetdb':
 					querycmd = 'mclient -h 127.0.0.1 -p 50001 -fcsv -s "%s" > /dev/null' % query
-					jdbcflags = '-u jdbc:monetdb://127.0.0.1:50001/database -n monetdb -p monetdb -d nl.cwi.monetdb.jdbc.MonetDriver'
+					jdbcflags = 'nl.cwi.monetdb.jdbc.MonetDriver jdbc:monetdb://127.0.0.1:50001/database monetdb monetdb'
 					odbccmd = 'isql MonetDB -d, < query > /dev/null'
 				elif system['db'] == 'db2':
 					db2qfile = open("db2query", "w")
 					db2qfile.write("connect to remotedb user user using user; \n" + query + ";\n")
 					db2qfile.close()
 					querycmd = 'db2 -tf db2query > /dev/null;'
-					jdbcflags = '-u jdbc:db2://127.0.0.1:50000/db -d com.ibm.db2.jcc.DB2Driver -n user -p user'
+					jdbcflags = 'com.ibm.db2.jcc.DB2Driver jdbc:db2://127.0.0.1:50000/db user user'
 					os.environ['DB2INSTANCE'] = 'user'
 					odbccmd = 'isql DB2_SAMPLE -d, user user < query > /dev/null'
 				elif system['db'] == 'oracle':
 					os.environ['TNS_ADMIN'] = '/home/user/oracleconfig'
 					querycmd = 'sqlplus system/oracle@//127.0.0.1:49161/XE @query-oracle.sql %d > /dev/null' % tuple
-					jdbcflags = '-u jdbc:oracle:thin:@127.0.0.1:49161:XE -d oracle.jdbc.driver.OracleDriver -n system -p oracle'
+					jdbcflags = 'oracle.jdbc.driver.OracleDriver jdbc:oracle:thin:@127.0.0.1:49161:XE system oracle'
 					odbccmd = 'isql Oracle -d, < query > /dev/null'
 					# for JDBC/ODBCV
 					query = "SELECT * FROM lineitem where rownum < %d;" % tuple
@@ -160,6 +161,11 @@ for r in range(nruns):
 					jdbcflags = None
 					odbccmd = None
 					odbcdriver = None
+				elif system['db'] == 'hive':
+					querycmd = None
+					jdbcflags = 'org.apache.hive.jdbc.HiveDriver jdbc:hive2://localhost:10000 user null'
+					odbccmd = None
+					odbcdriver = None
 				elif system['db'] == 'netcat':
 					# open netcat for listening in a separate process
 					filename = '/home/user/lineitem-%d.csv' % tuple
@@ -185,7 +191,12 @@ for r in range(nruns):
 				qfile.write("\n")
 				qfile.close()
 
-				jdbccmd = 'java -Xmx10G -Djava.security.egd=file:/dev/./urandom -Djava.ext.dirs=/home/user/java/ -jar /home/user/java/sqlline.jar --fastConnect=true --outputformat=csv  --isolation=TRANSACTION_SERIALIZABLE --silent=true --showHeader=false %s < query > /dev/null' % jdbcflags
+				jdbccmd = 'java -Xmx10G -Djava.security.egd=file:/dev/./urandom -cp /home/user/java/pmjc.jar:/home/user/java/db2jcc4.jar:/home/user/java/monetdb-jdbc-2.23.jar:/home/user/java/mariadb-java-client-1.4.6.jar:/home/user/java/ojdbc6_g.jar:/home/user/java/postgresql-9.4.1209.jar:/home/user/java/hive-jdbc-2.1.0-standalone.jar:/home/user/java/hadoop-common-2.6.4.jar %s "%s" 1000' % (jdbcflags, query)
+				
+				# special case for hive
+				if querycmd is None:
+					querycmd = jdbccmd
+					jdbccmd = None
 
 				# getting caches hot
 				benchmark_command(querycmd, system, 'native', network, tuple, r, True)