At least the content for src tar ball is wrong...
All the pathes are started with a prefix hbase-2.2.1, like this
hbase-2.2.1bin/
hbase-2.2.1conf/
Of course I do not think this is necessary to sink a RC but let's fix the
tar ball first?
Guanghao Zhang 于2019年8月15日周四 上午11:52写道：
> File
We have an issue https://issues.apache.org/jira/browse/HBASE-22862, Let's
discuss there.
Thanks.
On Sat, Aug 17, 2019 at 1:17 AM Stack wrote:
> Interesting. The Cells differ in sequence id. Would seem to imply race
> around the getting of sequenceid writing the WAL. The crashes are happening
>
Interesting. The Cells differ in sequence id. Would seem to imply race
around the getting of sequenceid writing the WAL. The crashes are happening
on regionserver replay of WAL files recovering crashed server? Or when is
it happening? If the former, you might have to find the offending WAL and
