While it may be useful to have a scheduler for simple cases, I think there are too many variables to make it useful for everyone's case. For example, I recently wrote a script that uses the data import handler api to get the status, kick off the import, etc. However, before allowing it to just kick off, I needed to query the database where the data was coming from to make sure it had finished it's daily load and then if it hadn't finished, wait for awhile to see if it would, then the script could do the load. After the load is finished it does another check to ensure the expected number of docs was actually loaded by Solr based on the data from the database.
If a scheduler were built into Solr it probably would only cover the simple case and for production you'd probably need to write your own scripts and use your own scheduler anyways to ensure the loads are starting/completing as expected. > On Sep 1, 2015, at 1:09 PM, William Bell <billnb...@gmail.com> wrote: > > We should add a simple scheduler in the UI. It is very useful. To schedule > various actions: > > - Full index > - Delta Index > - Replicate > > > > >> On Tue, Sep 1, 2015 at 12:41 PM, Shawn Heisey <apa...@elyograg.org> wrote: >> >>> On 9/1/2015 11:45 AM, Troy Edwards wrote: >>> My initial thought was to use scheduling built with DIH: >>> http://wiki.apache.org/solr/DataImportHandler#Scheduling >>> >>> But I think just a cron job should do the same for me. >> >> The dataimport scheduler does not exist in any Solr version. This is a >> proposed feature, with the enhancement issue open for more than four years: >> >> https://issues.apache.org/jira/browse/SOLR-2305 >> >> I have updated the wiki page to state the fact that the scheduler is a >> proposed improvement, not a usable feature. >> >> Thanks, >> Shawn > > > -- > Bill Bell > billnb...@gmail.com > cell 720-256-8076
