Hi Konstantin If I understand correctly, coordinator 2 doesn't have to wait for all 30 days data to be available, just start with whatever available ones within past 30 days. I can only come up with naïve approach now, which is running coordinator 2 as time-based, and in workflow, create java(or shell) action to grab list of available input data within 30 days, and pass to next action using <capture-output />, and next action consume that as input, and do actual processing.
Hope it helps Ryota On 4/4/13 3:49 AM, "Konstantin Zaytsev" <[email protected]> wrote: >Hi everyone! >How do i implement the following datasets synchronization scenario: >coordinator1 produces daily datasets and may sometimes fail to produce one >coordinator2 consumes all available datasets for the past 30 days. > >Both coord:current() and coord:latest() don't seem to be able to do it - >current() causes coordinator2 to hang in case some daily dataest is >missing, latest() just takes last 30 available (some of which could be >older than 30 days). > >Every suggestion is welcome and appreciated)
