| Addshore added a subscriber: Smalyshev. Addshore added a comment. |
In T204469#4588389, @Jonas wrote:> Propose
BDD GIVEN a query service related constraint check is run AND the query service sends a 429 response AND a "Retry-after" header is present THEN QualityConstraints should disable using the query service for that number of seconds
Yup, that BDD looks good.
I tailed off writing my version of the BDD as I found the open question about if the Retry-After is always provided by WDQS, as with a 429 response I don't believe it is actually required.
@Smalyshev could you let us know if a Retry-After will always be provided?If disabled AFAIK it falls back to php implementation and does not check all format and some of the (value-)type constraints.
Basically like if there was no service URL defined.Sounds good.
Is there any indication in the resulting constraint report stating that not all constraints have been run or something similar?@Lucas_Werkmeister_WMDE @Ladsgroup Maybe we could use some existing LockManager functionality that we already have?
Lockmanager probably isn't the right thing to use, I dont think you can set TTLs on locks in lock manager. You have to actually unlock them.
In T204469#4588156, @Lucas_Werkmeister_WMDE wrote:Tricky to do, because we need to persist the information about the rate limit somewhere between requests. Any ideas for that?
Would could store the timestamp of when the next request should occur in redis or memcached.
Then before the php code goes to run constraint reports using sparql it cn check for the key and see if it is allowed to run the query or not.
Cc: Smalyshev, Jonas, Ladsgroup, Lucas_Werkmeister_WMDE, Addshore, Aklapper, Lahi, Gq86, GoranSMilovanovic, QZanden, LawExplorer, Agabi10, Wikidata-bugs, aude, Lydia_Pintscher, Mbch331
_______________________________________________ Wikidata-bugs mailing list Wikidata-bugs@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikidata-bugs
