Description
When we set the start key and the end key in the function:
createTable(HTableDescriptor desc, byte[] startKey, byte[] endKey, int numRegions)
The current pre-split algorithm could not find a split point between the keys like "aaa" and "aab", "1111" and "1112".
Example Code for this bug:
admin.createTable(htd, Bytes.toBytes("aaa"), Bytes.toBytes("aab"), 4);
we will get the following ERROR:
Exception in thread "main" java.lang.IllegalArgumentException: Unable to split key range into enough regions
at org.apache.hadoop.hbase.client.HBaseAdmin.createTable(HBaseAdmin.java:473)
at test.JavaTest.main(JavaTest.java:28)
We hope this pre-split algorithm should be able to calculate the split point with an additional byte. for example:
"aaa" and "aab", split point= "aaaP"
"1111" and "1112", split point ="1111P"
Attachments
Issue Links
- duplicates
-
HBASE-12716 A bug in RegionSplitter.UniformSplit algorithm
- Closed