Clojurians Log v2
datahike
Is there any way to do a lazy pull over a large query result? like qseq in datomic? Or do I just need to build a composite tuple of the key I want to index on and use datoms?
If so, is there a way to make it so that a tuple will not insert if one of the elements is nil?
I'm getting an odd error. I'm doing serialized synchronous transactions on a datahike-jdbc connection backed by sqlite, and at some point an exception is being thrown by sqlite because it's complaining of being locked. This is in and of itself interesting and probably warrants I go look in on my jdbc connection or something, but it's not my primary issue. My primary issue is that this error results in the writer thread for the connection shutting down, and then all subsequent attempts I make at creating a new connection to the same sqlite database have transactions hang forever, even though queries work fine.
Regarding the error that the queries hang, you can release the connection completely https://cljdoc.org/d/io.replikativ/datahike/0.6.1572/api/datahike.api#release, by calling (release conn true) and then reconnect.
Thanks! I'll give that a shot to see if it resolves my issue when I create the minimal reproduction case.
That's still on my todo list to get done, I've just been busy
No problem. The locking error definitely should not happen.
Hey @suskeyhose! Do you happen to have the error at hand? It would be good to open a github issue to keep track of everything. This definitely should not happen.
It is a bit strange that there is a locking issue as the database should not be opened and closed while the datahike connection is open and I thought sqlite only locks at these points, but I guess it doesn't.
Also if you have a minimal reproducible case at hand and can add that that would be of great help!
I'll work on producing one!
Thank you!