The latest zotero beta 9 takes a long time to open

llity

Is there a download link for the previous version? thank you
https://s3.amazonaws.com/zotero.org/images/forums/u4956245/r2asquhr7mvs4r8wsbmo.png

llity

It has been resolved. Thank you

llity

itemDataValues is Zotero's value deduplication table, storing unique copies of all field values (titles, summaries, DOIs, etc.). The problem with Zotero is that when an entry is deleted or modified, the old value is not cleared from this table, resulting in orphan rows constantly piling up.

1,534 documents normally only require about 5 MB, but after long-term use (addition, deletion, modification), the orphan rows have accumulated to 17 GB.


$ sqlite3 ~/Zotero/zotero.sqlite "
SELECT name, SUM(pgsize) / 1024 / 1024 AS size_mb
FROM dbstat
GROUP BY name
ORDER BY size_mb DESC
LIMIT 10;
"
sqlite_autoindex_itemDataValues_1|9476
itemDataValues|7756
syncCache|7
itemNotes|2
sqlite_autoindex_fulltextWords_1|1
fulltextWords|1
version|0
users|0
translatorCache|0
tags|0

dstillman

Something is very wrong with your DB if you're seeing those numbers.

The problem with Zotero is that when an entry is deleted or modified, the old value is not cleared from this table, resulting in orphan rows constantly piling up.

No, this is false. (Please don't try to make authoritative claims about Zotero if you don't actually know how it works.) Zotero purges orphaned itemDataValues rows at startup, trash emptying, and sync time after you've deleted items.

If you run VACUUM; on the DB, what size is it after that?

dstillman

@llity: Can you also run these commands and let us know the output?

PRAGMA auto_vacuum;

SELECT COUNT(*) FROM items;

SELECT COUNT(*) FROM itemDataValues WHERE valueID NOT IN (SELECT valueID FROM itemData);