WikiData Import 2022-06-24
Wolfgang Fahl
Contents
Target:QLever
✅ This indexing attempt was successful.
see QLever/script as discussed in QLever Issue #562 for the script which makes reproducing this attempt easier.
see QLever Discussions for more details on this attempt series.
since the https://github.com/ad-freiburg/qlever-control now has an official "qlever" script we have rename the script that has the purpose to make the import attempts reproducible to qleverauto.
Beware of https://github.com/ad-freiburg/qlever-control/issues/4 - make sure ulimit -n is set!. This attempt had to be restarted since setting the value within a script did not work.
Preparations
Target:QLever[edit]
✅ This indexing attempt was successful.
see QLever/script as discussed in QLever Issue #562 for the script which makes reproducing this attempt easier.
see QLever Discussions for more details on this attempt series.
since the https://github.com/ad-freiburg/qlever-control now has an official "qlever" script we have rename the script that has the purpose to make the import attempts reproducible to qleverauto.
Beware of https://github.com/ad-freiburg/qlever-control/issues/4 - make sure ulimit -n is set!. This attempt had to be restarted since setting the value within a script did not work.
Contents
Preparations[edit]
Native build[edit]
WikiData_Import_2022-05-21#Build_code steps still apply for this attempt using the native/compiled version of qlever.
Native build[edit]
WikiData_Import_2022-05-21#Build_code steps still apply for this attempt using the native/compiled version of qlever.
Update qlever-code[edit]
qlever-code$ git pull
remote: Enumerating objects: 301, done.
remote: Counting objects: 100% (301/301), done.
remote: Compressing objects: 100% (223/223), done.
Receiving objects: 31% (94/301), 53.42 MiB | 3.75 MiB/s
...
create mode 100644 src/util/antlr/ANTLRErrorHandling.h
delete mode 100644 src/util/antlr/ThrowingErrorStrategy.h
create mode 100644 toolchains/gcc12.cmake
Update submodules[edit]
git submodule update --init --recursive
Submodule path 'third_party/abseil-cpp': checked out '2617970857c46e6ec971865d54f00445c260f682'
Submodule path 'third_party/googletest': checked out '0320f517fd920866d918e564105d68fd4362040a'
From https://github.com/ad-freiburg/stxxl
* branch 70cc597f3f76f96f036db4ffdd84a5cd7b224c7c -> FETCH_HEAD
Fetching submodule extlib/foxxll
Submodule path 'third_party/stxxl': checked out '70cc597f3f76f96f036db4ffdd84a5cd7b224c7c'
From https://github.com/ad-freiburg/foxxll
* branch 784859bc09a3982d6545fbf1d7b698e273401703 -> FETCH_HEAD
Submodule path 'third_party/stxxl/extlib/foxxll': checked out '784859bc09a3982d6545fbf1d7b698e273401703'
Build[edit]
wf@sun:/hd/seel/qlever/qlever-code$ rm -rf build/
wf@sun:/hd/seel/qlever/qlever-code$ mkdir build
wf@sun:/hd/seel/qlever/qlever-code$ cd build
wf@sun:/hd/seel/qlever/qlever-code/build$ cmake -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_COMPILER="g++-11" -DLOGLEVEL=INFO -DUSE_PARALLEL=true -GNinja ..
-- The C compiler identification is GNU 9.4.0
-- The CXX compiler identification is GNU 11.1.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
...
-- ---
-- Configuring done
-- Generating done
-- Build files have been written to: /hd/seel/qlever/qlever-code/build
Ninja[edit]
see https://ninja-build.org/manual.html
ninja
...
[613/613] Linking CXX executable test/SparqlExpressionTest
qleverauto environment checks[edit]
./qleverauto -v
qleverauto version : 1.29 $ : 2022/05/23 06:15:28 $
./qleverauto -e
needed software
docker → /usr/bin/docker ✅
top → /usr/bin/top ✅
df → /usr/bin/df ✅
jq → /usr/bin/jq ✅
lsb_release → /usr/bin/lsb_release ✅
free → /usr/bin/free ✅
operating system
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 20.04.4 LTS
Release: 20.04
Codename: focal
docker version
Docker version 20.10.16, build aa7e414
memory
total used free shared buff/cache available
Mem: 125Gi 1,3Gi 120Gi 27Mi 4,2Gi 123Gi
Swap: 2,0Gi 0B 2,0Gi
diskspace
/dev/sdb5 116G 25G 86G 23% /
tmpfs 63G 16K 63G 1% /dev/shm
/dev/sda1 3,6T 2,7T 716G 80% /hd/seel
/dev/sdb1 511M 4,0K 511M 1% /boot/efi
soft ulimit for files
1048576
wikidata files and index settings[edit]
reuse files of latest attempt of last month
wikidata$ ls -l
total 92754792
-rw-rw-r-- 1 wf wf 94653250500 Mai 19 07:35 latest-all.ttl.bz2
-rw-rw-r-- 1 wf wf 327629685 Mai 21 01:28 latest-lexemes.ttl.bz2
-rw-r--r-- 1 wf wf 911 Mai 22 17:47 Qleverfile
drwxrwxr-x 2 wf wf 4096 Mai 23 08:07 RCS
-rw-rw-r-- 1 wf wf 40 Mai 23 14:52 wikidata.settings.json
Qleverfile[edit]
cat Qleverfile
# Qleverfile for folder /hd/seel/qlever
# Automatically created on Sa 21. Mai 09:09:41 CEST 2022.
# Modify or expand as you see fit.
# Indexer settings
DB = wikidata
RDF_FILES = "latest-all.ttl.bz2 latest-lexemes.ttl.bz2"
CAT_FILES = "bzcat ${RDF_FILES}"
WITH_TEXT = false
SETTINGS_JSON = '{ "languages-internal": ["en"], "prefixes-external": [ "<http://www.wikidata.org/entity/statement", "<http://www.wikidata.org/value", "<http://www.wikidata.org/reference" ], "locale": { "language": "en", "country": "US", "ignore-punctuation": true }, "ascii-prefixes-only": true, "num-triples-per-batch": 10000000 }'
# Server settings
HOSTNAME = sun.bitplan.com
SERVER_PORT = 7001
MEMORY_FOR_QUERIES = 10
CACHE_MAX_SIZE_GB = 5
CACHE_MAX_SIZE_GB_SINGLE_ENTRY = 1
CACHE_MAX_NUM_ENTRIES = 100
# QLever binaries
QLEVER_BIN_DIR = /hd/seel/qlever/qlever-code/build/
USE_DOCKER = false
QLEVER_DOCKER_IMAGE = adfreiburg/qlever
QLEVER_DOCKER_CONTAINER = qlever.must_specify
# QLever UI
QLEVERUI_PORT = 7000
QLEVERUI_DIR = qlever-ui
QLEVERUI_CONFIG = default
rcsdiff -r1.2 ./Qleverfile
===================================================================
RCS file: ./RCS/Qleverfile,v
retrieving revision 1.2
diff -r1.2 ./Qleverfile
10,11c10
< SETTINGS_JSON = '{ "num-triples-per-batch": 10000000 }'
<
---
> SETTINGS_JSON = '{ "languages-internal": ["en"], "prefixes-external": [ "<http://www.wikidata.org/entity/statement", "<http://www.wikidata.org/value", "<http://www.wikidata.org/reference" ], "locale": { "language": "en", "country": "US", "ignore-punctuation": true }, "ascii-prefixes-only": true, "num-triples-per-batch": 10000000 }'
wikidata.settings.json[edit]
cat wikidata.settings.json
{
"languages-internal": ["en"],
"prefixes-external": [
"<http://www.wikidata.org/entity/statement",
"<http://www.wikidata.org/value",
"<http://www.wikidata.org/reference"
],
"locale": {
"language": "en",
"country": "US",
"ignore-punctuation": true
},
"ascii-prefixes-only": true,
"num-triples-per-batch" : 10000000
}
rcsdiff ./wikidata.settings.json
===================================================================
RCS file: ./RCS/wikidata.settings.json,v
retrieving revision 1.1
diff -r1.1 ./wikidata.settings.json
2c2,14
< "num-triples-per-batch": 10000000
---
> "languages-internal": ["en"],
> "prefixes-external": [
> "<http://www.wikidata.org/entity/statement",
> "<http://www.wikidata.org/value",
> "<http://www.wikidata.org/reference"
> ],
> "locale": {
> "language": "en",
> "country": "US",
> "ignore-punctuation": true
> },
> "ascii-prefixes-only": true,
> "num-triples-per-batch" : 10000000
Contents
Indexer[edit]
Relevant part of qleverauto script[edit]
#
# build the wikidata index
#
wikidata_index() {
cd $QLEVER_HOME/wikidata
chmod o+w .
show_timing "creating wikidata index" "started"
# docker run -i --rm -v $QLEVER_HOME/qlever-indices/wikidata:/index --entrypoint bash $dockerimage -c "cd /index && bzcat latest-all.ttl.bz2 latest-lexemes.ttl.bz2 | IndexBuilderMain -F ttl -f - -l -i wikidata -s wikidata.settings.json | tee wikidata.index-log.txt"
. ../qlever-control/qlever
check_installed IndexBuilderMain
qlever index
show_timing "creating wikidata index" "finished"
}
Relevant part of qleverauto script[edit]
#
# build the wikidata index
#
wikidata_index() {
cd $QLEVER_HOME/wikidata
chmod o+w .
show_timing "creating wikidata index" "started"
# docker run -i --rm -v $QLEVER_HOME/qlever-indices/wikidata:/index --entrypoint bash $dockerimage -c "cd /index && bzcat latest-all.ttl.bz2 latest-lexemes.ttl.bz2 | IndexBuilderMain -F ttl -f - -l -i wikidata -s wikidata.settings.json | tee wikidata.index-log.txt"
. ../qlever-control/qlever
check_installed IndexBuilderMain
qlever index
show_timing "creating wikidata index" "finished"
}
Symbolic Link for IndexBuilder Main[edit]
~/bin$ ls -l IndexBuilderMain
lrwxrwxrwx 1 wf wf 50 Mai 22 17:47 IndexBuilderMain -> /hd/seel/qlever/qlever-code/build/IndexBuilderMain
which IndexBuilderMain
/home/wf/bin/IndexBuilderMain
start[edit]
ulimit -n 1000000
ulimit -a | grep '(-n)'
open files (-n) 1000000
nohup ./qleverauto -wi&
tail -f nohup.out
bzcat latest-all.ttl.bz2 latest-lexemes.ttl.bz2 | IndexBuilderMain -F ttl -K wikidata -f - -i wikidata -s wikidata.settings.json | tee wikidata.index-log.txt
2022-06-24 10:17:24.091 - INFO: QLever IndexBuilder, compiled on Jun 24 2022 09:48:21
2022-06-24 10:17:24.092 - INFO: You specified the input format: TTL
2022-06-24 10:17:24.092 - INFO: You specified "locale = en_US" and "ignore-punctuation = 1"
2022-06-24 10:17:24.092 - INFO: You specified "ascii-prefixes-only = true", which enables faster parsing for well-behaved TTL files
2022-06-24 10:17:24.092 - INFO: You specified "num-triples-per-batch = 10,000,000", choose a lower value if the index builder runs out of memory
2022-06-24 10:17:24.092 - INFO: Integers that cannot be represented by QLever will throw an exception (this is the default behavior)
2022-06-24 10:17:24.092 - INFO: Processing input triples from /dev/stdin ...
2022-06-24 10:19:42.628 - INFO: Input triples processed: 100,000,000
2022-06-24 10:21:49.591 - INFO: Input triples processed: 200,000,000
...
2022-06-24 16:25:43.584 - INFO: Input triples processed: 17,300,000,000
2022-06-24 16:27:12.709 - INFO: Input triples processed: 17,400,000,000
2022-06-24 16:28:13.800 - INFO: Done, total number of triples read: 17,460,734,729 [may contain duplicates]
2022-06-24 16:28:13.800 - INFO: Number of QLever-internal triples created: 10,625,008,810 [may contain duplicates]
2022-06-24 16:28:13.800 - INFO: Merging partial vocabularies in byte order (internal only) ...
2022-06-24 16:30:11.559 - INFO: Words merged: 100,000,000
2022-06-24 16:32:56.452 - INFO: Words merged: 200,000,000
...
2022-06-24 16:50:30.791 - INFO: Words merged: 800,000,000
2022-06-24 16:50:32.869 - INFO: Number of words in internal vocabulary: 803,012,009
2022-06-24 16:50:32.870 - INFO: Merging partial vocabularies in Unicode order (internal and external) ...
2022-06-24 16:55:19.386 - INFO: Words merged: 100,000,000
2022-06-24 17:00:53.965 - INFO: Words merged: 200,000,000
...
2022-06-24 19:22:05.305 - INFO: Words merged: 3,000,000,000
2022-06-24 19:25:42.043 - INFO: Words merged: 3,100,000,000
2022-06-24 19:30:22.289 - INFO: Number of words in external vocabulary: 2,380,290,824
2022-06-24 19:30:22.289 - INFO: Removing temporary files ...
2022-06-24 19:30:36.250 - INFO: Converting external vocabulary to binary format ...
2022-06-24 20:06:10.166 - INFO: Converting triples from local IDs to global IDs ...
2022-06-24 20:06:40.823 - INFO: Triples converted: 100,000,000
2022-06-24 20:07:05.558 - INFO: Triples converted: 200,000,000
...
2022-06-24 22:06:47.931 - INFO: Triples converted: 28,000,000,000
2022-06-24 22:07:11.040 - INFO: Done, total number of triples converted: 28,085,743,539
2022-06-24 22:07:11.108 - INFO: Building prefix tree from internal vocabulary ...
2022-06-24 22:07:31.625 - INFO: Words processed: 100,000,000
2022-06-24 22:08:08.792 - INFO: Words processed: 200,000,000
...
2022-06-24 22:12:21.901 - INFO: Words processed: 800,000,000
2022-06-24 22:12:23.131 - INFO: Computing maximally compressing prefixes (greedy algorithm) ...
2022-06-24 22:29:23.381 - INFO: Reduction of size of internal vocabulary: 46%
2022-06-24 22:29:51.140 - INFO: Writing compressed vocabulary to disk ...
2022-06-25 00:34:02.750 - INFO: Creating a pair of index permutations ...
2022-06-25 04:21:33.258 - INFO: Statistics for PSO: #relations = 62,405, #blocks = 46,498, #triples = 23,411,086,943
2022-06-25 04:21:33.258 - INFO: Statistics for POS: #relations = 62,405, #blocks = 46,498, #triples = 23,411,086,943
2022-06-25 04:21:33.258 - INFO: Exchanging multiplicities for PSO and POS ...
2022-06-25 04:21:33.264 - INFO: Writing meta data for PSO and POS ...
2022-06-25 05:43:51.603 - INFO: Creating a pair of index permutations ...
2022-06-25 07:36:56.397 - INFO: Statistics for SPO: #relations = 2,779,853,585, #blocks = 29,774, #triples = 23,411,086,943
2022-06-25 07:36:56.397 - INFO: Statistics for SOP: #relations = 2,779,853,585, #blocks = 29,774, #triples = 23,411,086,943
2022-06-25 07:36:56.397 - INFO: Exchanging multiplicities for SPO and SOP ...
2022-06-25 08:18:51.323 - INFO: Writing meta data for SPO and SOP ...
2022-06-25 08:19:08.547 - INFO: Number of distinct patterns: 7,087,677
2022-06-25 08:19:08.547 - INFO: Number of subjects with pattern: 2,779,853,585 [all]
2022-06-25 08:19:08.547 - INFO: Total number of distinct subject-predicate pairs: 11,003,206,558
2022-06-25 08:19:08.547 - INFO: Average number of predicates per subject: 4.0
2022-06-25 08:19:08.580 - INFO: Average number of subjects per predicate: 236,837
2022-06-25 09:42:57.517 - INFO: Creating a pair of index permutations ...
creating wikidata index finished at Sa 25. Jun 11:04:33 CEST 2022 after 89230 seconds
Resulting files[edit]
-rwx------ 1 wf staff 94653250500 19 Mai 07:35 latest-all.ttl.bz2
-rwx------ 1 wf staff 327629685 21 Mai 01:28 latest-lexemes.ttl.bz2
-rwx------ 1 wf staff 1772814860288 25 Jun 09:42 wikidata-stxxl.disk
-rwx------ 1 wf staff 37505 25 Jun 09:42 wikidata.index-log.txt
-rwx------ 1 wf staff 28278087680 25 Jun 11:04 wikidata.index.ops
-rwx------ 1 wf staff 48982650912 25 Jun 11:04 wikidata.index.ops.meta
-rwx------ 1 wf staff 27706970112 25 Jun 11:04 wikidata.index.osp
-rwx------ 1 wf staff 48982650912 25 Jun 11:04 wikidata.index.osp.meta
-rwx------ 1 wf staff 573472800 25 Jun 11:04 wikidata.index.osp.tmp.mmap-buffer
-rwx------ 1 wf staff 15051429124 25 Jun 08:19 wikidata.index.patterns
-rwx------ 1 wf staff 56401897399 25 Jun 04:21 wikidata.index.pos
-rwx------ 1 wf staff 65137493514 25 Jun 04:21 wikidata.index.pso
-rwx------ 1 wf staff 36301340744 25 Jun 08:18 wikidata.index.sop
-rwx------ 1 wf staff 110210977824 25 Jun 08:18 wikidata.index.sop.meta
-rwx------ 1 wf staff 36699637749 25 Jun 08:18 wikidata.index.spo
-rwx------ 1 wf staff 110210977824 25 Jun 08:18 wikidata.index.spo.meta
-rwx------ 1 wf staff 274 24 Jun 22:40 wikidata.meta-data.json
-rwx------ 1 wf staff 4295 24 Jun 22:29 wikidata.prefixes
-rwx------ 1 wf staff 362 24 Jun 10:17 wikidata.settings.json
-rwx------ 1 wf staff 155452489143 24 Jun 20:06 wikidata.vocabulary.external
-rwx------ 1 wf staff 48982650912 24 Jun 20:06 wikidata.vocabulary.external.idsAndOffsets.mmap
-rwx------ 1 wf staff 27373372956 24 Jun 22:39 wikidata.vocabulary.internal
Memory usage[edit]
2022-06-25 09:30
top
...
MiB Mem : 128832,8 total, 5514,4 free, 12913,2 used, 110405,2 buff/cache
Starting server[edit]
ServerMain -i /hd/seel/qlever/wikidata/wikidata -p 7001
2022-08-10 16:41:45.203 - INFO: Done, number of words: 808,217,822
2022-08-10 16:41:45.206 - INFO: Number of words in external vocabulary: 2,406,607,301
2022-08-10 16:41:45.235 - INFO: Registered PSO permutation: #relations = 63,016, #blocks = 46,898, #triples = 23,623,846,556
2022-08-10 16:41:45.267 - INFO: Registered POS permutation: #relations = 63,016, #blocks = 46,898, #triples = 23,623,846,556
2022-08-10 16:41:45.276 - INFO: Registered OPS permutation: #relations = 3,179,186,092, #blocks = 39,688, #triples = 23,623,846,556
2022-08-10 16:41:45.286 - INFO: Registered OSP permutation: #relations = 3,179,186,092, #blocks = 39,688, #triples = 23,623,846,556
2022-08-10 16:41:45.293 - INFO: Registered SPO permutation: #relations = 2,808,258,381, #blocks = 30,045, #triples = 23,623,846,556
2022-08-10 16:41:45.302 - INFO: Registered SOP permutation: #relations = 2,808,258,381, #blocks = 30,045, #triples = 23,623,846,556
2022-08-10 16:41:45.302 - INFO: Reading patterns from file /hd/seel/qlever/wikidata/wikidata.index.patterns ...
2022-08-10 16:42:26.209 - INFO: Sorting random result tables to estimate the sorting performance of this machine ...
2022-08-10 16:42:43.558 - INFO: Access token for restricted API calls is ""
2022-08-10 16:42:43.558 - INFO: The server is ready, listening for requests on port 7001 ...
WikiData Import 2022-06-24 Wolfgang Fahl