Allow user-defined batch size in ColumnSliceIterator #422

Closed
patricioe opened this Issue Feb 17, 2012 · 2 comments

Comments

Projects
None yet
2 participants
Collaborator

patricioe commented Feb 17, 2012

No description provided.

patricioe was assigned Feb 17, 2012

Collaborator

patricioe commented Feb 17, 2012

The default batch size (aka count) is set to 100. For some applications this default is too small.

For instance, reading 33k small columns with default count = 100 takes 10x more time than reading in batches of 10k.

Thus, let's the client define its own batch size if needed.

@patricioe patricioe added a commit that referenced this issue Feb 17, 2012

@patricioe patricioe Allows the client specify the batch size #422
It applies to ColumnSliceIterator
61e577a

patricioe closed this Feb 17, 2012

orrgal1 commented Aug 23, 2012

just used it with hector 1.1 and at least for small numbers it doesnt work. i have a CF with 5 members. even if i put 2 in the count i still get all 5.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment