[
https://issues.apache.org/jira/browse/TEPHRA-214?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15855516#comment-15855516
]
ASF GitHub Bot commented on TEPHRA-214:
---------------------------------------
Github user poornachandra commented on a diff in the pull request:
https://github.com/apache/incubator-tephra/pull/31#discussion_r99760732
--- Diff:
tephra-hbase-compat-1.1-base/src/main/java/org/apache/tephra/hbase/txprune/DataJanitorState.java
---
@@ -129,17 +168,19 @@ public long getPruneUpperBoundForRegion(byte[]
regionId) throws IOException {
Result next;
while ((next = scanner.next()) != null) {
byte[] region = getRegionFromKey(next.getRow());
- if (regions.contains(region)) {
- byte[] timeBytes = next.getValue(FAMILY,
PRUNE_UPPER_BOUND_COL);
- if (timeBytes != null) {
- long pruneUpperBoundRegion = Bytes.toLong(timeBytes);
- resultMap.put(region, pruneUpperBoundRegion);
+ if (regions == null || regions.contains(region)) {
+ Cell cell = next.getColumnLatestCell(FAMILY,
PRUNE_UPPER_BOUND_COL);
+ if (cell != null) {
+ byte[] pruneUpperBoundBytes = CellUtil.cloneValue(cell);
+ long timestamp = cell.getTimestamp();
--- End diff --
This is not compaction timestamp, this is when the prune upper bound was
written to HBase. Although this can be an approximate value when the state
table is available. Is the compaction timestamp really necessary for debugging?
> Provide a way to debug Transaction Pruning progress
> ---------------------------------------------------
>
> Key: TEPHRA-214
> URL: https://issues.apache.org/jira/browse/TEPHRA-214
> Project: Tephra
> Issue Type: Improvement
> Reporter: Gokul Gunasekaran
> Assignee: Gokul Gunasekaran
>
> It would be good to get information about what regions are lagging that might
> be useful to debug why transaction pruning might not be happening. We do
> print debug information but this method could be invoked on a need-basis
> outside of the regular pruning operation.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
