On 04/09/11 07:43, Mathias Herberts wrote:
On Sep 4, 2011 1:39 AM, "Bill de hÓra"<[email protected]> wrote:
On 02/09/11 19:06, Stack wrote:
What do folks think?
Not putting the log into hdfs seems like a good idea.
I was somehow thinking the opposite as it makes irrecoverable machine
failures much more problematic. What makes you say it's a good idea?
Grants control over the write path, specifically sequential I/O and
crash recovery. Granted you'd want more than one copy replicated, but
you need that regardless. Thinking a bit more it might not work with the
regionserver model, and the Accumulo proposal mentions a service here
rather than a local disk. Anyway, I may be still thinking about older
HBase code when the WAL didn't work so well, today it seems like just
edge cases (https://issues.apache.org/jira/browse/HBASE-4107).
Bill
