Our application will be pretty much Write Once/Read Many, with rare
updates to any object. We'll have < 500TiB of total storage with < 1B
objects. While I'm leaning toward a single bucket to hold the data, I do
have some questions:
Is there an advantage to sharding the data being as it will be
exclusively read?
What are the practical limits on the number of objects in one bucket?
We probably will want to get the list of objects in the bucket(s) -
should that factor into the sharding/single bucket decision?
Thanks,
Mark
--
Mark Winnemueller
reThought DevOps Engineer
719.480.4609
mark.winnemueller@xxxxxxxxxxxxxxxxxxxxxx
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
