It works, it has worked in the past... One thing to remember is you can't always split... You cant split until all the open scanners are closed. Furthermore you can't split if a region has just split. The reason for this is because once you split a region, files aren't rewritten, but the 2 daughter regions hold references to the parent region. Until those regions are resolved via compaction (slow, minutes possibly), you can't split those daughter regions.
So check again, maybe your splits are working after all. If you are doing imports, check out my Randomize.java i posted to the list last week. That helps import performance substantially. On Wed, Apr 29, 2009 at 12:52 AM, Edward J. Yoon <[email protected]>wrote: > Hi, > > I saw that function on the web UI but it seems not implemented yet. > > public void modifyTable(final byte[] tableName, int op, Writable[] args) > switch (op) { > ... > case MODIFY_TABLE_SPLIT: > case MODIFY_TABLE_COMPACT: > .. > } > } > > Am I right? Pls let me know if i'm wrong. > -- > Best Regards, Edward J. Yoon @ NHN, corp. > [email protected] > http://blog.udanax.org >
