Addshore added a comment.

  In T241212#5784995 <>, 
@Ladsgroup wrote:
  > Fun thing is that we basically can't be 100% sure if any row is orphan 
unless we lock both tables and won't let any write queries happening on them 
which is not possible but given that these tables are huge and queries to find 
orphan rows would long time we can't say for sure there has been writes in the 
mean time making us think some rows are orphan (the easiest way to mitigate 
most of the issue is to trimming all rows that have a high PK id
  Could have a look at the hadoop snapshots (as they dont get updated live ;))



To: Ladsgroup, Addshore
Cc: Ladsgroup, Addshore, Aklapper, Iflorez, darthmon_wmde, alaa_wmde, Nandana, 
Lahi, Gq86, GoranSMilovanovic, QZanden, LawExplorer, _jensen, rosalieper, 
Scott_WUaS, Jonas, Wikidata-bugs, aude, Lydia_Pintscher, Mbch331
Wikidata-bugs mailing list

Reply via email to