get_indexed_slices timing out in Pycassa, indexes working in CLI #96

Closed
buddhasystem opened this Issue Nov 11, 2011 · 4 comments

3 participants

@buddhasystem

I have a rather straightforward index (on of many), which is DATE. When I query by date in cassandra cli, the result comes back instantaneously. However, when I'm doing same in a pycassa client, even if I put in a limit on only 100 items, the application times out on the pool. This is identical for 1.0.8 and 1.3.0. Thank you, Maxim

@buddhasystem

Partially resolved, but I suspect that iteration of an indexed query result can silently fail not.

@thobbs
pycassa member

What do you mean by "I suspect that iteration of an indexed query result can silently fail not"? Can you provide more details?

@buddhasystem

Hello Tyler,
sorry for a garbled post. What I observed was this: I do a query using a secondary index, in a few batches. I delete rows based on what I find, and do the next query with the same criterion. This query gets stuck after a few iterations. Thank you.

@christopherkeller

Pretty sure I'm seeing the exact same issue, but for me it fails consistently. Works fine in the CLI and it works fine with one particular value of the index (which returns 65K rows). When I try the other value of the index via pycassa fails to return any of the ~250 rows, with a timeout exception. I believe the problem to be an underlying issue with Thrift, as we see similar errors using the Ruby libraries.

@thobbs thobbs closed this Jun 24, 2013
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment