lookup ColumnSliceIterator in hector. this does what you want a lot easier than managing the "paging"
From: Yuhan Zhang <yzh...@onescreen.com<mailto:yzh...@onescreen.com>> Reply-To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" <user@cassandra.apache.org<mailto:user@cassandra.apache.org>> Date: Tue, 14 Feb 2012 14:46:08 -0800 To: "user@cassandra.apache.org<mailto:user@cassandra.apache.org>" <user@cassandra.apache.org<mailto:user@cassandra.apache.org>> Subject: Re: timed-out retrieving a giant row. Thanks Robin! :) On Tue, Feb 14, 2012 at 12:23 PM, R. Verlangen <ro...@us2.nl<mailto:ro...@us2.nl>> wrote: Of course you should set your limit to 100 or something like that, not Integer.MAX_VALUE ;-) 2012/2/14 R. Verlangen <ro...@us2.nl<mailto:ro...@us2.nl>> I'm familiar to this in PHPCassa, but with Hector it would be something like this: Query you CF with a range.setStart(lastColName) and range.setFinish(StringUtils.byte("") where the " lastColName " is the name of the column from the previous read. You can continue this until you run out of results. 2012/2/14 Yuhan Zhang <yzh...@onescreen.com<mailto:yzh...@onescreen.com>> Hi all, I'm using the Hector client 0.8, trying to retrieve a list of IDs from a gaint row. each ID is a columnName in the row It works ok when there's not many IDs, but SliceQuery starts to time-out after the row becomes big. Is this approach the correct way to store a list of IDs? are there some settings that I'm missing? by looking at the code, it sets the range of the columnNames to be setRange(null, null, false, Integer.MAX_VALUE); is there a way in cassandra to retrieve the first 100 columns, then the next 100 columns, and so forth? Thank you. Yuhan
