Addshore added a comment.
replies to 2 comments inline In T221774#5296984 <https://phabricator.wikimedia.org/T221774#5296984>, @alaa_wmde wrote: > That thing is definitely better to live outside of Wikibase codebase, if possible, in an extension/external library that is more Wikidata. is there such place? what is "Wikidata.org extension"? and did you want to name it `WikidataPrometheusSparqlEndpointReplicationStatus` (Wikidata instead of Wikimedia), or why Wikimedia? The Wikidata.org extension might make sense. BUT Once SDOC and commons starts sending data to the wdqs then we will also need this code to be on commons. the Wikidata.org extension is only on wikidata.org. There is also the WikimediaEvents extension, that by name this doesnt totally fit into, but might be the "right" place for now, and conditionally run for wikibases hooked up to a query service. > I'm also wondering if the whole thing (interface and other implmenetation `SparqlEndpointReplicationStatusStateHandler` etc) can as well live outside Wikibase Repo? (I'm assuming Wikibase Repo has no code related to WDQS but I can certainly be off here, haven't checked yet). > >> Do we (also?) want to have a pure SPARQL implementation? > > In case we do Prometheus based solution, then I wouldn't keep two implementations unless I want to use one as a fallback for the other one. Are you thinking of using SPARQL implementation as a fallback in case Prometheus is not reachable? Thinking about this more I think this code could live just fine in Wikibase.get itself, and just be configurable. There is no reason that over users of wikibase and the query service would not want to use this. In T221774#5297123 <https://phabricator.wikimedia.org/T221774#5297123>, @hoo wrote: > We could create a small hook in Wikibase that lets us inject arbitrary things into max lag… and then in Wikidata.org <https://github.com/wikimedia/mediawiki-extensions-Wikidata.org> do all the interaction with Prometheus and inject the lag. See comment above about wikidata.org vs wikibase repose in general. >>> Do we (also?) want to have a pure SPARQL implementation? >> >> In case we do Prometheus based solution, then I wouldn't keep two implementations unless I want to use one as a fallback for the other one. Are you thinking of using SPARQL implementation as a fallback in case Prometheus is not reachable? > > I mostly thought that might be useful for third party instance that only run one-instance query services... but it might be better to stick with YAGNI and just implement this in Wikidata.org for Wikidata. I agree, not yet :) lets just fix it for us first. TASK DETAIL https://phabricator.wikimedia.org/T221774 EMAIL PREFERENCES https://phabricator.wikimedia.org/settings/panel/emailpreferences/ To: hoo, Addshore Cc: alaa_wmde, Andrawaag, Multichill, Ladsgroup, Smalyshev, fgiunchedi, hoo, Daniel_Mietchen, MisterSynergy, Addshore, Sjoerddebruin, Aklapper, Lucas_Werkmeister_WMDE, Daryl-TTMG, RomaAmorRoma, 0010318400, E.S.A-Sheild, darthmon_wmde, joker88john, CucyNoiD, Nandana, NebulousIris, Gaboe420, Versusxo, Majesticalreaper22, Giuliamocci, Adrian1985, Cpaulf30, Lahi, Gq86, Af420, Darkminds3113, Bsandipan, Lordiis, GoranSMilovanovic, Adik2382, Chicocvenancio, Th3d3v1ls, Ramalepe, Liugev6, QZanden, EBjune, merbst, LawExplorer, WSH1906, Lewizho99, Volans, Maathavan, _jensen, rosalieper, Jonas, Xmlizer, jkroll, Wikidata-bugs, Jdouglas, aude, Tobias1984, Manybubbles, Lydia_Pintscher, Mbch331
_______________________________________________ Wikidata-bugs mailing list [email protected] https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs
