Research talk:Quarry

From Meta, a Wikimedia project coordination wiki

Page categories[edit]

Should this page be categorized in Category:Research? --Bamyers99 (talk) 20:49, 10 February 2015 (UTC)Reply[reply]

Execution time[edit]

Hi, is there any limit in the execution time of a query? If not, is it somehow possible to trigger a long-running query, close the browser window and later download the result? Thanks, --Aschroet (talk) 21:09, 15 October 2015 (UTC)Reply[reply]

Hi Aschroet. Right now, there's a 20 minute execution time. You can start a query, close the browser and expect that the query will continue to run. --EpochFail (talk) 21:13, 15 October 2015 (UTC)Reply[reply]
I don't know why, but i just got a result of a query that ran over night with leaving the browser open. How can i know the run_number when i start the query to get the result later? --Aschroet (talk) 03:51, 16 October 2015 (UTC)Reply[reply]
How do you need to get the result? Are you running a script which uses the result? The run number is given in the HTML source code of the query page. wget -O - http://quarry.wmflabs.org/run/`wget -O - http://quarry.wmflabs.org/query/nnnn | grep 'qrun_id' | sed 's/^.*qrun_id\": \([^,]*\),.*$/\1/g'`/output/0/tsv?download=true will send the query result set to the standard output on UNIX systems, if you replace nnnn with the query number. --Stefan2 (talk) 10:05, 16 October 2015 (UTC)Reply[reply]

Why not run queries in parallel?[edit]

Today I experienced my queries running slowly or being queued. Why can't we have multiple queries in parallel, as it will help users to not wait for any trouble some queries of others to complete/timeout.Example query which might be causing my queries to wait. http://quarry.wmflabs.org/query/6185 . I found this query running for more than one hour.--Arjunaraoc (talk) 10:02, 27 March 2016 (UTC)Reply[reply]

Hello Arjunaraoc, there are two worker servers that process each one multiple queries at the same time. The problem you describe should no longer be present, fell free to ping me if it's re-appear. Thanks --Framawiki (talk) 17:32, 8 June 2018 (UTC)Reply[reply]

Help with query[edit]

I'm trying to make a query that list of all "ns 0" articles (not disambigs or redirects) with 0 incoming links (alernativly 0 incoming "ns 0" links).

https://quarry.wmflabs.org/query/9103 is getting killed for taking too long time. Anyone able to help me optimize this, or perhaps run it on PAWS or similar? Josve05a (talk) 22:27, 10 May 2016 (UTC)Reply[reply]

mw:Talk:Quarry is generally better for asking for help.
Try using cl_to IN ('All_set_index_articles', 'All_disambiguation_pages') instead of using the page_props table. That might speed up the query a bit. --Stefan2 (talk) 22:45, 10 May 2016 (UTC)Reply[reply]

Tool labs' user databases?[edit]

Is it possible to query tools' user databases? For example, the ToolTranslate tools' database s53069__tooltranslate_p? (And thanks for a brilliant thing; Quarry is great.) — Sam Wilson ( TalkContribs ) … 00:42, 27 July 2016 (UTC)Reply[reply]

Hello Sam Wilson, it's two different database servers, it should not be possible actually. It's phab:T151158. --Framawiki (talk) 17:32, 8 June 2018 (UTC)Reply[reply]

How long do results lag behind live data?[edit]

Sorry if I've missed this, but it would be helpful if this page of documentation indicated how much time lag there is of the replicas being queried behind the live data. (I'm querying the recent_changes table in wikidatawiki_p.) Mhl20 (talk) 12:25, 9 May 2018 (UTC)Reply[reply]

Hello Mhl20, it's phab:T60841. In the meanwhile you can use https://tools.wmflabs.org/replag/ as indicator. --Framawiki (talk) 17:32, 8 June 2018 (UTC)Reply[reply]

Unable to log in[edit]

I have been unable to log in to Quarry. Even I permit to access with OAuth, this message "Internal Server Error The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application." was displayed. Is it caused by my wrong? Please help me.--Miraburu (talk) 07:44, 1 May 2019 (UTC)Reply[reply]

Running a query in multiple wikis[edit]

Is there a convenient way to run the same query in multiple languages?

Or do I have to change the "use xxwiki_p" line at the top and rerun it for every wiki? --Amir E. Aharoni (talk) 14:04, 5 December 2019 (UTC)Reply[reply]

Trying to ping: User:Zhuyifei1999, User:Framawiki. --Amir E. Aharoni (talk) 07:36, 12 December 2019 (UTC)Reply[reply]
Not that I am aware of :( --Zhuyifei1999 (talk) 15:18, 12 December 2019 (UTC)Reply[reply]
Not yet possible. That's phab:T95582. One day(TM) --Framawiki (talk) 18:04, 12 December 2019 (UTC)Reply[reply]
Heh, a bug I reported by myself and forgot :)
Thanks Framawiki! --Amir E. Aharoni (talk) 11:42, 17 December 2019 (UTC)Reply[reply]

Delete past query[edit]

Hello, is there anyway I can delete my past queries? I have several drafts piling up now and would like to clean it while a few are concerning privacy. RXerself (talk) 06:04, 5 May 2021 (UTC)Reply[reply]

RXerself, you could overwrite them? https://quarry.wmflabs.org/query/21660 - Cabayi (talk) 14:54, 6 May 2021 (UTC)Reply[reply]

Global edits[edit]

I don't know much about SQL, but I managed to scrape some code together that allowed me to get the top editors on my home wiki by edit count, but I was wondering if there is way to do this for global edits. – Ilovemydoodle (talk) 07:53, 4 September 2022 (UTC)Reply[reply]

Global edit count is stored in a separate table global_edit_count, but it is apparently not available in public replicas and it probaly wouldn't useful on its own either because 1) the edit count is not indexed, 2) the referenced table is not available, too. --Matěj Suchánek (talk) 09:57, 4 September 2022 (UTC)Reply[reply]

MediaWiki plugin QueryViz and Quarry ?[edit]

Hello Quarry team,
I'm one of the maintainer of Lingualibre.org (a Wikimedia tool)'s Mediawiki pluging QueryViz. It currently works with Wikibase endpoints : switch to the right endpoint, POST the query, receive the (json) response, build and inject the results as a ikitable into the wikipage. We recently added Wikidata SPARQL endpoint successfully. Commons SPARQL endpoints being private, our xhr requests on it fail. I wonder if :

  • Is it possible to xhr POST crossdomain queries from https://lingualibre.org to quarry.wmcloud.org successfully ?
  • Will you attend the 2023 Wikimedia Hackathon in Athens ? Where we could meet.

Yug (talk) 18:57, 11 April 2023 (UTC)Reply[reply]