Mercurial > hg > cc > work

changeset 84:5aa8b75aa9f0

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Confirmed new approach matches ks{,_errs} for the three cases we have: 0, 4 and 15
author Henry S. Thompson <ht@inf.ed.ac.uk>
date Mon, 14 Apr 2025 15:22:32 +0100
parents d41cef3d4582
children 19e2136f595b
files lurid3/notes.txt
diffstat 1 files changed, 359 insertions(+), 1 deletions(-) [+]
line wrap: on
line diff
--- a/lurid3/notes.txt	Wed Apr 09 21:53:55 2025 +0100
+++ b/lurid3/notes.txt	Mon Apr 14 15:22:32 2025 +0100
@@ -1837,7 +1837,7 @@
 
 OK, rerun in the right place:
   >: cd ~/results/CC-MAIN-2023-40/warc_lmhx
-  >: time python3 -c 'import sys,warc2cdb; sys.exit(warc2cdb.main(*sys.argv[1:]))'  2023-40 15 . 2> 15/w2c_errs &/work/dc007/dc007/hst/lib/python/cc:/work/dc007/dc007/hst/lib/python/cc/lmh
+  >: time python3 -c 'import sys,warc2cdb; sys.exit(warc2cdb.main(*sys.argv[1:]))'  2023-40 15 . 2> 15/w2c_errs &
 
   real    210m54.183s
   user    152m34.172s
@@ -2384,7 +2384,365 @@
   x4/11/w2c_errs:180
   >: tail -1 slurm_*11n*
   Wed Apr 9 21:39:53 BST 2025 Finished
+
 Need some sanity checks before gearing up production...
+
+ >: head -1402 11/w2c_errs | diff -bw - <(fgrep -v records x1/11/w2c_errs)
+ [nothing]
+
+Argh.  Missing a few hits at the end of each file:
+  >: python3 -m w2c 2023-40 11 plus 000,002 2> plus/11/w2c_err
+  >: grep records plus/11/w2c_errs
+  115308 records, max record: 1058293, max header: 20289
+  114429 records, max record: 1058293, max header: 20289
+  114729 records, max record: 1058293, max header: 20289
+  >: uz /beegfs/common_crawl/CC-MAIN-2023-40/1695233506559.11/orig/warc/CC-MAIN-20230924023050-20230924053050-00000.warc.gz |egrep -ac '^WARC-Type: '
+  115408
+
+Last record from each buffer-load?  No, the entire last buffer-full.
+Fixed.
+
+  >: python3 -m w2c 2023-40 11 plus 000,002 2> plus/11/w2c_errs
+  plus/11/lmh.cdb_in
+  >: grep records plus/11/w2c_errs
+  115408 records, max record: 1058293, max header: 20289
+  114508 records, max record: 1058293, max header: 20289
+  114865 records, max record: 1058293, max header: 20289
+  >: head -12 11/w2c_errs |diff -bw - plus/11/w2c_errs
+  2a3,4
+  > 115408 records, max record: 1058293, max header: 20289
+  > 6186 LM headers, 6185 win, 1 lose, 0 non-ASCII URIs
+  4a7,8
+  > 114508 records, max record: 1058293, max header: 20289
+  > 6149 LM headers, 6148 win, 1 lose, 1 non-ASCII URIs
+  12c16,17
+  < /beegfs/common_crawl/CC-MAIN-2023-40/1695233506559.11/orig/warc/CC-MAIN-20230924023050-20230924053050-00003.warc.gz
+  ---
+  > 114865 records, max record: 1058293, max header: 20289
+  > 6192 LM headers, 6186 win, 6 lose, 1 non-ASCII URIs
+
+Retry the 5-by test:
+
+  >: fgrep -c beegfs x?/11/w*
+  x0/11/w2c_errs:180
+  x1/11/w2c_errs:180
+  x2/11/w2c_errs:180
+  x3/11/w2c_errs:180
+  x4/11/w2c_errs:180
+  >: head -1 slurm_*11n*
+  Thu Apr 10 15:35:12 BST 2025 Launching 3 nodes ...
+  >: tail -1 slurm_*11n*
+  Thu Apr 10 16:35:48 BST 2025 Finished
+
+A minute or two slower.
+
+Error logs still look OK:
+
+ >: head -1402 11/w2c_errs | diff -bw - <(fgrep -v records x1/11/w2c_errs)
+ [nothing]
+
+Right number of records for 00000
+  >: grep records x1/11/w2c_errs |head -3
+  115408 records, max record: 1058293, max header: 20289
+  114508 records, max record: 1058293, max header: 20289
+  114865 records, max record: 1058293, max header: 20289
+  >: grep records x1/11/w2c_errs |tail -1
+  115216 records, max record: 1079547, max header: 40116
+  >: for i in 0 2 3 4; do grep records x$i/11/w2c_errs |tail -1 ; done
+  115654 records, max record: 1077742, max header: 40071
+  115978 records, max record: 1077845, max header: 16008
+  115483 records, max record: 1079052, max header: 25052
+  114757 records, max record: 1079623, max header: 40083
+  >: diff -bw 11/lmh.cdb_in x1/11/lmh.cdb_in |egrep -v '^<'
+  1130232,1174166c1130232
+  >: for i in {0..4}; do grep 'LM headers' x$i/11/w2c_errs|cut -f 4 -d ' ' |btot ; done| cut -f 4 -d ' ' |tee >(cat 1>&2) | btot
+  1132831
+  1130231
+  1131595
+  1131756
+  1131479
+  5657892
+
+Belt and braces -- rerun all the existing full 900 cases except 11
+  >: for d in 0 15 1 surplus/4; do mkdir -p tentative/$d; mv $d/{w2c_errs,lmh.cdb_in} tentative/$d; done
+  >: for d in 0; do time ~/lib/python/cc/w2c.py 2023-40 $d . 2> $d/w2c_errs; done &
+  Thu Apr 10 10:50:37 PM BST 2025
+
+  real    279m24.973s
+  user    176m3.292s
+  sys     18m21.907s
+  ./0/lmh.cdb_in
+
+  >: date; for d in  1 2 4 15; do { ~/lib/python/cc/w2c.py 2023-40 $d . 2> $d/w2c_errs ; echo $d finished $(date); } & done
+  ./1/lmh.cdb_in
+  1 finished Fri Apr 11 03:30:01 AM BST 2025
+  ./4/lmh.cdb_in
+  4 finished Fri Apr 11 03:31:44 AM BST 2025
+  ./2/lmh.cdb_in
+  2 finished Fri Apr 11 03:35:17 AM BST 2025
+  ./15/lmh.cdb_in
+  15 finished Fri Apr 11 03:38:49 AM BST 2025
+
+  >: cut -f 2 <(sed 's/value or format .*/value or format/' 0/w2c_errs)|egrep -v 'warc.gz|LM headers|records,'| sus
+     4381 Invalid date value or format
+	1 year 4263994 is out of range
+	1 year 4278897 is out of range
+	1 year 4355304 is out of range
+	1 year 4361928 is out of range
+	1 year 4370444 is out of range
+	1 year 4430056 is out of range
+	1 year 4433368 is out of range
+	1 year 53310 is out of range
+  >: cut -f 2 0/ks_errs.tsv | sus
+     3188 cannot unpack non-iterable NoneType object
+     1193 list index out of range
+	1 year 4263994 is out of range
+	1 year 4278897 is out of range
+	1 year 4355304 is out of range
+	1 year 4361928 is out of range
+	1 year 4370444 is out of range
+	1 year 4430056 is out of range
+	1 year 4433368 is out of range
+	1 year 53310 is out of range
+Same for 4 and 15
+
+  >: cd 0
+  >: wc -l ks.tsv
+  6414380 ks.tsv
+  >: wc -l lmh.cdb_in
+  6414381 lmh.cdb_in
+  >: diff -bw <(cut -f2- ks.tsv| sed 's/ //;s/\.0$//'|sort -k1,1) <(sed 's/^\+[0-9,]*://;s/->\([0-9]*\)$/ \1/' lmh.cdb_in | sort -k1,1)|head
+
+  >: ls -l {tentative/,}{0,4,15}/lmh.cdb_in tentative/surplus/4/lmh.cdb_in 
+  -rw-r--r-- 1 hst dc007 719413017 Apr 11 03:24 0/lmh.cdb_in
+  -rw-r--r-- 1 hst dc007 719413017 Mar 24 16:52 tentative/0/lmh.cdb_in
+  -rw-r--r-- 1 hst dc007 603244719 Apr 11 03:38 15/lmh.cdb_in
+  -rw-r--r-- 1 hst dc007 603244719 Mar 19 22:56 tentative/15/lmh.cdb_in
+  -rw-r--r-- 1 hst dc007 704148582 Apr 11 03:31 4/lmh.cdb_in
+  -rw-r--r-- 1 hst dc007 704148582 Mar 27 02:12 tentative/surplus/4/lmh.cdb_in
+
+But:
+
+  >: diff -bw <(cut -f2- ks.tsv| sed 's/ //;s/\.0$//'|sort -k1,1) <(sed 's/^\+[0-9,]*://;s/->\([0-9]*\)$/ \1/' lmh.cdb_in | sort -k1,1)|head
+  14634c14635
+  < 20230930082240https://www.levnepletaciprize.cz/shop/p\u0159\u00EDze-pro-miminka/elen-baby?error=404    1696062160
+  ---
+  > 20230930082240https://www.levnepletaciprize.cz/shop/p%C5%99%C3%ADze-pro-miminka/elen-baby?error=404 1696062160
+  25306c25307
+  < 20230930082317https://zums.ac.ir/content/68251/\u0647\u0632\u06CC\u0646\u0647-\u0627\u0639\u0645\u0627\u0644-\u062C\u0631\u0627\u062D\u06CC-\u06F1\u06F4\u06F0\u06F1    1696058653
+  ---
+  > 20230930082317https://zums.ac.ir/content/68251/%D9%87%D8%B2%DB%8C%D9%86%D9%87-%D8%A7%D8%B9%D9%85%D8%A7%D9%84-%D8%AC%D8%B1%D8%A7%D8%AD%DB%8C-%DB%B1%DB%B4%DB%B0%DB%B1 1696058653
+
+Oops!
+
+  : fgrep -c \\u0 */ks.tsv
+  0/ks.tsv:1062
+  12/ks.tsv:679
+  15/ks.tsv:634
+  36/ks.tsv:1311
+  4/ks.tsv:770
+  56/ks.tsv:1194
+  68/ks.tsv:846
+  best_two_by_nl1/ks.tsv:1873
+
+Even worse:
+  >: cd ../../CC-MAIN-2019-35/warc_lmhx
+  >: fgrep -c \\u0 ?/ks.tsv
+  0/ks.tsv:3
+  1/ks.tsv:4
+  2/ks.tsv:4
+  3/ks.tsv:2
+  4/ks.tsv:0
+  5/ks.tsv:2
+  6/ks.tsv:1
+  7/ks.tsv:1
+  8/ks.tsv:8
+  9/ks.tsv:4
+
+Or maybe not?
+  >: fgrep \\u0 cluster.idx
+  at,wko,news)/news/system/400.html?error=400&errorrequest=/pdf/gen?url=https:/news.wko.at/news/oesterreich/wirtschaftsbarometer_wirtschaftslage.html?shorturl=wirtschaftsbarometerat&key=98c90396b77ff930d6df30ce0f7f7d9c&pdfoptions={"header-html":"\/news\/templates\/header_html.php","footer-html":"https:\/\/news.wko.at\/news\/templates\/footer_html.php?url=https://news.wko.at/news/oesterreich/wirtschaftsbarometer_wirtschaftslage.html?shorturl=wirtschaftsbarometerat","print-media-type":"","title":"wirtschaftsbarometer:+wk\u00d6-konjunkturumfrage"} 20190826093246   cdx-00001.gz     849440852       286441  8844
+  >: fgrep -c \\u0 /beegfs/common_crawl/CC-MAIN-2019-35/cdx/cluster.idx
+  1
+  sing<3771>: fgrep \\u0 /beegfs/common_crawl/CC-MAIN-2019-35/cdx/cluster.idx
+  at,wko,news)/news/system/400.html?error=400&errorrequest=/pdf/gen?url=https:/news.wko.at/news/oesterreich/wirtschaftsbarometer_wirtschaftslage.html?shorturl=wirtschaftsbarometerat&key=98c90396b77ff930d6df30ce0f7f7d9c&pdfoptions={"header-html":"\/news\/templates\/header_html.php","footer-html":"https:\/\/news.wko.at\/news\/templates\/footer_html.php?url=https://news.wko.at/news/oesterreich/wirtschaftsbarometer_wirtschaftslage.html?shorturl=wirtschaftsbarometerat","print-media-type":"","title":"wirtschaftsbarometer:+wk\u00d6-konjunkturumfrage"} 20190826093246   cdx-00001.gz     844682060       287090  8844
+
+Try using the codecs hack from sort_date.py
+
+What's the best segment to test on...
+
+  >: pwd
+  /beegfs/common_crawl/CC-MAIN-2023-40/cdx/warc
+  >: uz cdx-00001.gz |egrep -a '\\u[0-9a-z][0-9a-z]'|egrep -o '"filename": "crawl-data.*' | cut -f 4 -d / | cut -f 2 -d . | sus
+       79 91
+       72 80
+       63 52
+       62 78
+       42 56
+       40 15
+       ...
+
+No, those are mostly (all for seg 15) crawl-diagnostics
+Ah, but, sigh, c-d/.../warc.gz do _have_ timestamps
+
+>: uz cdx-0001?.gz |egrep -a '\\u[0-9a-z][0-9a-z]'|egrep -o '"filename": "crawl-data.*/warc/' | cut -f 4 -d / | cut -f 2 -d . | sus
+      9 57
+      6 97
+      4 49
+      4 6
+      3 11
+      ...
+
+  >: for f in cdx-0001?.gz; do echo -n "$f "; egrep -a '\\u[0-9a-z][0-9a-z]' <(uz $f)|egrep '"filename": "crawl-data.*/warc/' | fgrep -c '.11/warc'; done
+  cdx-00010.gz 0
+  ...
+  cdx-00017.gz 3
+
+  >: for f in cdx-00017.gz; do egrep -a '\\u[0-9a-z][0-9a-z]' <(uz $f)|egrep '"filename": "crawl-data.*/warc/' | fgrep '.11/warc'; done
+  cn,kbstore,auth)/index.php/apireguser/flag/8?code=jdtong&code=jdtong&data={"logo":"http:\/\/jdtong.kbstore.cn\/public\/images\/logo-1.png","logo_url":"http:\/\/jdtong.kbstore.cn\/admin?flag=8","title":"\u8857\u5e97\u901a" [...] "filename": "crawl-data/CC-MAIN-2023-40/segments/1695233506559.11/warc/CC-MAIN-20230924023050-20230924053050-00003.warc.gz", "charset": "UTF-8", "languages": "zho"}
+  cn,kbstore,auth)/index.php/apireguser/flag/8?code=jdtong&code=jdtong&data={"logo":"http:\/\/jdtong.kbstore.cn\/public\/images\/logo-1.png","logo_url":"http:\/\/jiediantong.cn\/admin?flag=8","title":"\u8857\u5e97\u901a" ... "filename": "crawl-data/CC-MAIN-2023-40/segments/1695233506559.11/warc/CC-MAIN-20230924023050-20230924053050-00225.warc.gz", "charset": "UTF-8", "languages": "zho"}
+  cn,kbstore,auth)/index.php/apireguser/flag/8?code=kbcms&code=kbcms&data={"logo":"http:\/\/demo.cms.vip100.ltd\/public\/images\/logo.png","logo_url":"http:\/\/demo.cms.vip100.ltd\/admin?flag=8","title":"\u4eba\u4eba\u7ad9cms" [...] "filename": "crawl-data/CC-MAIN-2023-40/segments/1695233506559.11/warc/CC-MAIN-20230924023050-20230924053050-00537.warc.gz", "charset": "UTF-8", "languages": "zho"}
+
+Red herring chase: none of those have timestamps.
+
+So how can this be true:
+  >: egrep -c ' [^0] non-ASCII' {0,1,2,4,15}/w2*
+  0/w2c_errs:774
+  1/w2c_errs:683
+  2/w2c_errs:685
+  4/w2c_errs:731
+  15/w2c_errs:643
+
+Because e.g. Korean appears non-uniformly in index files:
+
+  >: egrep '^kr,' cluster.idx |head -3
+  kr,07e,m)/service/agreement 20230924093051      cdx-00221.gz    816253375       185571   1005693
+  kr,114n,bit)/sys/login.php?amp;auto_no=568&returnurl=/sys/login.php?auto_no=568&returnurl=http://bit.114n.kr/sys/board.php?bo_id=m71 20230923204454      cdx-00221.gz     816438946       182780  1005694
+  kr,1bang)/posts/online-shopping/13076 20231003101132    cdx-00221.gz    816621726184260  1005695
+
+  >: uz cdx-00221.gz | egrep -c '\\u[0-9a-f]'
+  9575
+  >: uz cdx-00010.gz | egrep -c '\\u[0-9a-f]'
+  746
+
+  >: ~/lib/python/cc/w2c.py 2023-40 15 . 000 |& less
+  >: diff -bw 15/lmh.cdb_in ~/results/CC-MAIN-2023-40/warc_lmhx/15/lmh.cdb_in
+  98c98
+  < +99,10:20230922035528http://cn.coii.kr/board/\u5546\u54C1\u8D2D\u4E70\u540E\u8BB0/4/?link_product_no=19597->1695354928
+  ---
+  > +117,10:20230922035528http://cn.coii.kr/board/%E5%95%86%E5%93%81%E8%B4%AD%E4%B9%B0%E5%90%8E%E8%AE%B0/4/?link_product_no=19597->1695354928
+
+Try the whole thing:
+
+  >: sbatch --output=slurm_w2c_15f.out --time=02:00:00 --ntasks=5 -c 2 $HOME/bin/runme.sh $PWD -d -t 2 -i 'cat 11n.pats |share_by_task.sh $n $task' 'export PYTHONPATH=$W/hst/lib/python/cc:$W/hst/lib/python/cc/lmh
+  mkdir -p ./fixu/x$task/15
+  ~/lib/python/cc/w2c.py 2023-40 15 ./fixu/x$task "$arg"  2> ./fixu/x$task/15/w2c_errs'
+  Sun Apr 13 03:45:46 PM BST 2025 end 4
+  Sun Apr 13 03:45:46 PM BST 2025 task 5.4 on r1i0n20:2.1 end
+  ./fixu/x3/15/lmh.cdb_in
+  Sun Apr 13 03:45:50 PM BST 2025 end 3
+  Sun Apr 13 03:45:50 PM BST 2025 task 5.3 on r1i0n20:2.1 end
+  ./fixu/x2/15/lmh.cdb_in
+  Sun Apr 13 03:46:04 PM BST 2025 end 2
+  Sun Apr 13 03:46:04 PM BST 2025 task 5.2 on r1i0n20:2.1 end
+  ./fixu/x1/15/lmh.cdb_in
+  Sun Apr 13 03:46:41 PM BST 2025 end 1
+  Sun Apr 13 03:46:41 PM BST 2025 task 5.1 on r1i0n3:2.0 end
+  ./fixu/x0/15/lmh.cdb_in
+  Sun Apr 13 03:52:49 PM BST 2025 end 0
+  Sun Apr 13 03:52:49 PM BST 2025 task 5.0 on r1i0n3:2.0 end
+  Sun Apr 13 15:52:49 BST 2025 Finished
+  >: head -1 slurm_w2c_15f.out 
+  Sun Apr 13 14:59:54 BST 2025 Launching 2 nodes in /mnt/lustre/e1000/home/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx for runme /work/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx -d -t 2 -i cat 11n.pats |share_by_task.sh $n $task export PYTHONPATH=$W/hst/lib/python/cc:$W/hst/lib/python/cc/lmh
+
+~50:10 elapsed: Faster!
+
+  >: diff -bw <(cut -f2- 15/ks.tsv| sed 's/      //;s/\.0$//'|sort -k1,1) <(sed 's/^\+[0-9,]*://;s/->\([0-9-]*\)$/ \1/' fixu/x{0..4}/15/lmh.cdb_in | sort -k1,1)| tee /tmp/hst/udiff | less
+  >: cat /tmp/hst/udiff
+  0a1,5
+  >
+  >
+  >
+  >
+  >
+
+And correct.
+
+Another one:
+
+  >: sbatch --output=slurm_w2c_4f.out --time=01:30:00 --ntasks=5 -c 2 $HOME/bin/runme.sh $PWD -d -t 2 -i 'cat 11n.pats |share_by_task.sh $n $task' 'export PYTHONPATH=$W/hst/lib/python/cc:$W/hst/lib/python/cc/lmh
+  mkdir -p ./fixu/x$task/4
+  ~/lib/python/cc/w2c.py 2023-40 4 ./fixu/x$task "$arg"  2> ./fixu/x$task/4/w2c_errs'
+  Submitted batch job 7381438
+  cirrus-login1<6096>: fgrep -c beegfs fixu/x?/4/w* | cut -f 2 -d : | btot
+  900
+  >: head -1 slurm_w2c_4f.out 
+  Mon Apr 14 08:55:08 BST 2025 Launching 3 nodes in /mnt/lustre/e1000/home/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx for runme /work/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx -d -t 2 -i cat 11n.pats |share_by_task.sh $n $task export PYTHONPATH=$W/hst/lib/python/cc:$W/hst/lib/python/cc/lmh
+  >: tail -1 slurm_w2c_4f.out
+  Mon Apr 14 09:41:51 BST 2025 Finished
+
+46:43 elapsed
+
+  >: diff -bw <(cut -f2- 4/ks.tsv| sed 's/       //;s/\.0$//'|sort -k1,1) <(sed 's/^\+[0-9,]*://;s/->\([0-9-]*\)$/ \1/' fixu/x{0..4}/4/lmh.cdb_in | sort -k1,1)| tee /tmp/hst/udiff | less
+  >: cat /tmp/hst/udiff
+  0a1,5
+  >
+  >
+  >
+  >
+  >
+
+And errors:
+  >: for d in 4 15; do ls -ld $d; cut -f 2 <(sed 's/value or format .*/value or format/' fixu/x?/$d/w2c_errs)|egrep -v 'warc.gz|LM headers|records,'| sus ; cut -f 2 $d/ks_errs.tsv | sus ; done
+  drwxr-sr-x 2 hst dc007 110592 Apr 10 22:50 4
+     6107 Invalid date value or format
+	8 year 539500 is out of range
+	7 year 539498 is out of range
+	2 year 55722 is out of range
+	1 year 53337 is out of range
+	1 year 539499 is out of range
+	1 year 54618 is out of range
+	1 year 54752 is out of range
+     4605 cannot unpack non-iterable NoneType object
+     1502 list index out of range
+	8 year 539500 is out of range
+	7 year 539498 is out of range
+	2 year 55722 is out of range
+	1 year 53337 is out of range
+	1 year 539499 is out of range
+	1 year 54618 is out of range
+	1 year 54752 is out of range
+  drwxr-sr-x 2 hst dc007 98304 Apr 10 22:50 15
+     8090 Invalid date value or format
+      113 year 641471 is out of range
+       20 year 641474 is out of range
+	9 hour must be in 0..23
+	9 year 642435 is out of range
+	3 year 642437 is out of range
+	1 year 4262575 is out of range
+	1 year 641769 is out of range
+	1 year 642434 is out of range
+     5420 cannot unpack non-iterable NoneType object
+     2670 list index out of range
+      113 year 641471 is out of range
+       20 year 641474 is out of range
+	9 hour must be in 0..23
+	9 year 642435 is out of range
+	3 year 642437 is out of range
+	1 year 4262575 is out of range
+	1 year 641769 is out of range
+	1 year 642434 is out of range
+
+Those were both with warc2cdb.py.  Try a 3rd run for timing after
+cythonize
+
+  >: head -1 slurm_w2c_0f.out 
+  Mon Apr 14 14:15:55 BST 2025 Launching 1 nodes in /mnt/lustre/e1000/home/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx for runme /work/dc007/dc007/hst/results/CC-MAIN-2023-40/warc_lmhx -d -t 2 -i cat 11n.pats |share_by_task.sh $n $task export PYTHONPATH=$W/hst/lib/python/cc:$W/hst/lib/python/cc/lmh
+  >: tail -1 slurm_w2c_0f.out 
+  Mon Apr 14 15:03:41 BST 2025 Finished
+
+47:46 elapsed, mid-day weekday, all on one node...
+Both cross-checks succeed
 ================
 
 Try it with the existing _per segment_ index we have for 2019-35