[
https://issues.apache.org/jira/browse/HIVE-21884?focusedWorklogId=329478&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-329478
]
ASF GitHub Bot logged work on HIVE-21884:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 16/Oct/19 23:15
Start Date: 16/Oct/19 23:15
Worklog Time Spent: 10m
Work Description: jcamachor commented on pull request #794: HIVE-21884
URL: https://github.com/apache/hive/pull/794#discussion_r335749992
##########
File path:
ql/src/java/org/apache/hadoop/hive/ql/schq/ScheduledQueryExecutionService.java
##########
@@ -0,0 +1,159 @@
+package org.apache.hadoop.hive.ql.schq;
+
+import java.io.Closeable;
+import java.io.IOException;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.TimeUnit;
+
+import org.apache.hadoop.hive.conf.HiveConf;
+import org.apache.hadoop.hive.metastore.api.QueryState;
+import org.apache.hadoop.hive.metastore.api.ScheduledQueryPollResponse;
+import org.apache.hadoop.hive.metastore.api.ScheduledQueryProgressInfo;
+import org.apache.hadoop.hive.ql.DriverFactory;
+import org.apache.hadoop.hive.ql.IDriver;
+import org.apache.hadoop.hive.ql.processors.CommandProcessorResponse;
+import org.apache.hadoop.hive.ql.session.SessionState;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import com.google.common.annotations.VisibleForTesting;
+import com.google.common.util.concurrent.ThreadFactoryBuilder;
+
+public class ScheduledQueryExecutionService implements Closeable {
+
+ private static final Logger LOG =
LoggerFactory.getLogger(ScheduledQueryExecutionService.class);
+
+ private ScheduledQueryExecutionContext context;
+ private ScheduledQueryExecutor worker;
+
+ public static ScheduledQueryExecutionService
startScheduledQueryExecutorService(HiveConf conf) {
+ MetastoreBasedScheduledQueryService qService = new
MetastoreBasedScheduledQueryService(conf);
+ ExecutorService executor =
+ Executors.newCachedThreadPool(
+ new
ThreadFactoryBuilder().setDaemon(true).setNameFormat("Scheduled Query Thread
%d").build());
+ ScheduledQueryExecutionContext ctx = new
ScheduledQueryExecutionContext(executor, conf, qService);
+ return new ScheduledQueryExecutionService(ctx);
+ }
+
+ public ScheduledQueryExecutionService(ScheduledQueryExecutionContext ctx) {
+ context = ctx;
+ ctx.executor.submit(worker = new ScheduledQueryExecutor());
+ ctx.executor.submit(new ProgressReporter());
+ }
+
+ class ScheduledQueryExecutor implements Runnable {
+
+ private ScheduledQueryProgressInfo info;
+
+ @Override
+ public void run() {
+ while (true) {
+ ScheduledQueryPollResponse q =
context.schedulerService.scheduledQueryPoll();
+ if (q.isSetExecutionId()) {
+ try{
+ processQuery(q);
+ } catch (Throwable t) {
+ LOG.error("Unexpected exception during scheduled query
processing", t);
+ }
+ } else {
+ try {
+ Thread.sleep(context.getIdleSleepTime());
+ } catch (InterruptedException e) {
+ Thread.currentThread().interrupt();
+ LOG.info("interrupted");
+ break;
+ }
+ }
+ }
+ }
+
+ QueryState lastState;
+ public synchronized void reportQueryProgress() {
+ if (info != null) {
+ if(lastState == QueryState.FINISHED || lastState ==
QueryState.ERRORED) {
+ return;
+ }
+ LOG.info("Reporting query progress of {} as {} err:{}",
info.getScheduledExecutionId(), info.getState(),
+ info.getErrorMessage());
+ context.schedulerService.scheduledQueryProgress(info);
+ lastState=info.getState();
+ }
+ }
+
+ private void processQuery(ScheduledQueryPollResponse q) {
Review comment:
We need to fix 2. One of the common use cases that we were discussing for
scheduled queries was incrementally refreshing contents in a table in a
transactional way, e.g., pulling data from Kafka. Currently, if query can be
assumed dead and then come back to life, we may basically end up inserting data
more than once in target object. You could just grab an exclusive execution
lock on the scheduled query identifier to prevent this from happening, we
already have mechanisms to do that (fwiw we have similar logic for incremental
materialized view rebuild).
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Issue Time Tracking
-------------------
Worklog Id: (was: 329478)
> Scheduled query support
> -----------------------
>
> Key: HIVE-21884
> URL: https://issues.apache.org/jira/browse/HIVE-21884
> Project: Hive
> Issue Type: Improvement
> Reporter: Zoltan Haindrich
> Assignee: Zoltan Haindrich
> Priority: Major
> Labels: pull-request-available
> Attachments: HIVE-21844.04.patch, HIVE-21844.05.patch,
> HIVE-21844.06.patch, HIVE-21844.07.patch, HIVE-21844.08.patch,
> HIVE-21844.09.patch, HIVE-21844.15.patch, HIVE-21844.19.patch,
> HIVE-21844.21.patch, HIVE-21884.01.patch, HIVE-21884.02.patch,
> HIVE-21884.03.patch, HIVE-21884.09.patch, HIVE-21884.10.patch,
> HIVE-21884.10.patch, HIVE-21884.11.patch, HIVE-21884.12.patch,
> HIVE-21884.13.patch, HIVE-21884.14.patch, HIVE-21884.14.patch,
> HIVE-21884.14.patch, HIVE-21884.16.patch, HIVE-21884.17.patch,
> HIVE-21884.18.patch, HIVE-21884.20.patch, Scheduled queries2.pdf
>
> Time Spent: 5h 40m
> Remaining Estimate: 0h
>
> design document:
> https://docs.google.com/document/d/1mJSFdJi_1cbxJTXC9QvGw2rQ3zzJkNfxOO6b5esmyCE/edit#
> in case the google doc is not reachable: [^Scheduled queries2.pdf]
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
