which algorithm did pmm use in query analytics?

which algorithm did pmm use in query analytics?
on pmm query analytics, query id is 16 characters, and pt-query-digest generates 32 characters checksum in history table.
how can i join query id between pmm and pt-query-digest?
i’m using pmm v1.15 and pt-query-digest v3.0.12,
thanks in advance.

Both programs use different fingerprint libraries so, they is no way to link query ID’s from query-digest and QAN.


thanks for you reply,
when i find a slow sql on a mysql server with QAN, i want to know its historical execution information which is stored in historical table saved by query-digest,
it’s unconvenient if i could not link query id from query-digest and QAN,
is there any plan to unify them with one algorithm?

Hi romber
At this time we do not have plans to unify the two products in terms of query ID hashes. We’re investing the bulk of our Development team efforts in PMM - I am interested to hear your use case - why do you want to join these two products? What do you gain? Thanks!

hi @Michael Coburn,
as i mentioned above,
query-digest could store historical slow log information to history/review tables,
so if query id’s are unified, when i see a slow log on pmm QAN, i could use it’s query id to select its historical execution information in history/review tables,
this may help me to find useful information.

if there is no plan, could you tell me which finger print libraries and hash function did QAN use?