[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15981833#comment-15981833
]
Sahil Takiar commented on HIVE-14864:
-
I updated the documentation for this here:
https://cwiki.apache.org/confluence/display/Hive/Configuration+Properties#ConfigurationProperties-QueryandDDLExecution
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Labels: TODOC2.2
> Fix For: 2.2.0
>
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.4.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15906158#comment-15906158
]
Lefty Leverenz commented on HIVE-14864:
---
Doc note: This adds *hive.exec.copyfile.maxnumfiles* to HiveConf.java and
corrects the description of *hive.exec.copyfile.maxsize* (added in 1.1.0 by
HIVE-8750 but not documented yet) so they need to be documented in the wiki.
* [Configuration Properties -- Query and DDL Execution |
https://cwiki.apache.org/confluence/display/Hive/Configuration+Properties#ConfigurationProperties-QueryandDDLExecution]
Added a TODOC2.2 label.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Labels: TODOC2.2
> Fix For: 2.2.0
>
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.4.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15903491#comment-15903491
]
Sergio Peña commented on HIVE-14864:
The patch looks good [~stakiar]
+1
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.4.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15902953#comment-15902953
]
Hive QA commented on HIVE-14864:
Here are the results of testing the latest attachment:
https://issues.apache.org/jira/secure/attachment/12856936/HIVE-14864.4.patch
{color:green}SUCCESS:{color} +1 due to 1 test(s) being added or modified.
{color:red}ERROR:{color} -1 due to 1 failed/errored test(s), 10337 tests
executed
*Failed tests:*
{noformat}
org.apache.hive.jdbc.TestMultiSessionsHS2WithLocalClusterSpark.testSparkQuery
(batchId=219)
{noformat}
Test results: https://builds.apache.org/job/PreCommit-HIVE-Build/4047/testReport
Console output: https://builds.apache.org/job/PreCommit-HIVE-Build/4047/console
Test logs: http://104.198.109.242/logs/PreCommit-HIVE-Build-4047/
Messages:
{noformat}
Executing org.apache.hive.ptest.execution.TestCheckPhase
Executing org.apache.hive.ptest.execution.PrepPhase
Executing org.apache.hive.ptest.execution.ExecutionPhase
Executing org.apache.hive.ptest.execution.ReportingPhase
Tests exited with: TestsFailedException: 1 tests failed
{noformat}
This message is automatically generated.
ATTACHMENT ID: 12856936 - PreCommit-HIVE-Build
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.4.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15902460#comment-15902460
]
Sahil Takiar commented on HIVE-14864:
-
[~spena] I worked on this some more, and think a unit test may be better suited
for this patch rather than a qtest. There are a number of different queries
that could invoke this method (e.g. IMPORT queries use this method too), and
more may be added in the future. I added some integration tests that run
against a mini HDFS cluster, and some unit tests that just rely on mocking.
[~ste...@apache.org] I agree, calling {{getContentSummary}} on S3 will be very
slow. I've thought about filing a JIRA for the optimization you mentioned a few
times, but never got around to doing it. Fortunately, this specific code won't
be hit for S3, only for HDFS.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.4.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15895818#comment-15895818
]
Steve Loughran commented on HIVE-14864:
---
{{FileSystem.getContentSummary()}} does a recursive treewalk, so is
pathologically bad on a blobstore which has to mock directories through many.,
many HTTP requests.
If you need to use it, could you actually supply a patch (+ FS contract tests)
for the method so that it uses listFiles(path, recursive=true)? That does the
same treewalk against HDFS, but blobstores can do it as an O(1) listing call
instead. If you can get that patch in, then enumerating the size of a blobstore
tree will be fast
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15892913#comment-15892913
]
Sergio Peña commented on HIVE-14864:
Didn't the mini HDFS accepts hdfs? Check all the encryption tests, such as
encryption_ctas.q. They usually check for the hdfs scheme.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15891510#comment-15891510
]
Sahil Takiar commented on HIVE-14864:
-
[~spena] am trying to write some qtests for this, do you know which CliDrivers
run against HDFS filesystems? When I run the {{TestCliDriver}} all the URIs
have schemes of either {{file}} or {{pfile}}. This code path only gets
triggered if the scheme is {{hdfs}}.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=1550#comment-1550
]
Sergio Peña commented on HIVE-14864:
Can we add q-tests for this new variable?
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15883969#comment-15883969
]
Hive QA commented on HIVE-14864:
Here are the results of testing the latest attachment:
https://issues.apache.org/jira/secure/attachment/12854547/HIVE-14864.3.patch
{color:red}ERROR:{color} -1 due to no test(s) being added or modified.
{color:red}ERROR:{color} -1 due to 4 failed/errored test(s), 10259 tests
executed
*Failed tests:*
{noformat}
TestDerbyConnector - did not produce a TEST-*.xml file (likely timed out)
(batchId=235)
org.apache.hadoop.hive.cli.TestMiniLlapLocalCliDriver.testCliDriver[vector_if_expr]
(batchId=140)
org.apache.hadoop.hive.cli.TestPerfCliDriver.testCliDriver[query14]
(batchId=223)
org.apache.hadoop.hive.cli.TestPerfCliDriver.testCliDriver[query23]
(batchId=223)
{noformat}
Test results: https://builds.apache.org/job/PreCommit-HIVE-Build/3768/testReport
Console output: https://builds.apache.org/job/PreCommit-HIVE-Build/3768/console
Test logs: http://104.198.109.242/logs/PreCommit-HIVE-Build-3768/
Messages:
{noformat}
Executing org.apache.hive.ptest.execution.TestCheckPhase
Executing org.apache.hive.ptest.execution.PrepPhase
Executing org.apache.hive.ptest.execution.ExecutionPhase
Executing org.apache.hive.ptest.execution.ReportingPhase
Tests exited with: TestsFailedException: 4 tests failed
{noformat}
This message is automatically generated.
ATTACHMENT ID: 12854547 - PreCommit-HIVE-Build
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch,
> HIVE-14864.3.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15582459#comment-15582459
]
Sergio Peña commented on HIVE-14864:
A couple of comments:
* Will srcFS.getContentSummary(src) cause extra time when source is on S3? If
so, maybe we want to put this line inside the if() statement.
* Once we're here, could you fix the message from HIVE_EXEC_COPYFILE_MAXSIZE to
say (in Bytes) instead of (in Mb) ?
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15577556#comment-15577556
]
Hive QA commented on HIVE-14864:
Here are the results of testing the latest attachment:
https://issues.apache.org/jira/secure/attachment/12833472/HIVE-14864.2.patch
{color:red}ERROR:{color} -1 due to no test(s) being added or modified.
{color:red}ERROR:{color} -1 due to 8 failed/errored test(s), 10564 tests
executed
*Failed tests:*
{noformat}
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[acid_globallimit]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[order_null]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[union_fast_stats]
org.apache.hadoop.hive.cli.TestHBaseCliDriver.testCliDriver[hbase_bulk]
org.apache.hive.beeline.TestBeelineArgParsing.testAddLocalJarWithoutAddDriverClazz[0]
org.apache.hive.beeline.TestBeelineArgParsing.testAddLocalJar[0]
org.apache.hive.beeline.TestBeelineArgParsing.testAddLocalJar[1]
org.apache.hive.jdbc.authorization.TestJdbcWithSQLAuthorization.testBlackListedUdfUsage
{noformat}
Test results: https://builds.apache.org/job/PreCommit-HIVE-Build/1581/testReport
Console output: https://builds.apache.org/job/PreCommit-HIVE-Build/1581/console
Test logs: http://104.198.109.242/logs/PreCommit-HIVE-Build-1581/
Messages:
{noformat}
Executing org.apache.hive.ptest.execution.TestCheckPhase
Executing org.apache.hive.ptest.execution.PrepPhase
Executing org.apache.hive.ptest.execution.ExecutionPhase
Executing org.apache.hive.ptest.execution.ReportingPhase
Tests exited with: TestsFailedException: 8 tests failed
{noformat}
This message is automatically generated.
ATTACHMENT ID: 12833472 - PreCommit-HIVE-Build
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Sahil Takiar
> Attachments: HIVE-14864.1.patch, HIVE-14864.2.patch, HIVE-14864.patch
>
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15573955#comment-15573955
]
Hive QA commented on HIVE-14864:
Here are the results of testing the latest attachment:
https://issues.apache.org/jira/secure/attachment/12833234/HIVE-14864.patch
{color:red}ERROR:{color} -1 due to no test(s) being added or modified.
{color:red}ERROR:{color} -1 due to 89 failed/errored test(s), 10534 tests
executed
*Failed tests:*
{noformat}
TestMiniLlapLocalCliDriver - did not produce a TEST-*.xml file
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[autoColumnStats_7]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[crtseltbl_serdeprops]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[cte_3]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[cte_4]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[cte_mat_4]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[cte_mat_5]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_00_nonpart_empty]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_01_nonpart]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_02_00_part_empty]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_02_part]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_03_nonpart_over_compat]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_04_all_part]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_04_evolved_parts]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_05_some_part]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_06_one_part]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_07_all_part_over_nonoverlap]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_08_nonpart_rename]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_09_part_spec_nonoverlap]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_10_external_managed]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_11_managed_external]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_12_external_location]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_13_managed_location]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_14_managed_location_over_existing]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_15_external_part]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_16_part_external]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_17_part_managed]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_18_part_external]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_19_00_part_external_location]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_19_part_external_location]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_20_part_managed_location]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_21_export_authsuccess]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_22_import_exist_authsuccess]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_23_import_part_authsuccess]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_24_import_nonexist_authsuccess]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_25_export_parentpath_has_inaccessible_children]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[exim_hidden_files]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[foldts]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[masking_9]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[materialized_view_describe]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[reloadJar]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[repl_2_exim_basic]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[repl_3_exim_metadata]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[stats_partial_size]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[temp_table]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[temp_table_gb1]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[temp_table_join1]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[temp_table_subquery1]
org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver[vector_tablesample_rows]
org.apache.hadoop.hive.cli.TestEncryptedHDFSCliDriver.testCliDriver[encryption_load_data_to_encrypted_tables]
org.apache.hadoop.hive.cli.TestMiniLlapCliDriver.testCliDriver[import_exported_table]
org.apache.hadoop.hive.cli.TestMiniSparkOnYarnCliDriver.testCliDriver[import_exported_table]
org.apache.hadoop.hive.cli.TestNegativeCliDriver.testCliDriver[authorization_import]
org.apache.hadoop.hive.cli.TestNegativeCliDriver.testCliDriver[ctas_noemptyfolder]
org.apache.hadoop.hive.cli.TestNegativeCliDriver.testCliDriver[exim_01_nonpart_over_loaded]
org.apache.hadoop.hive.cli.TestNegativeCliDriver.testCliDriver[exim_02_all_part_over_overlap]
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15536989#comment-15536989
]
Sergio Peña commented on HIVE-14864:
I run some tests with directories, and distcp is indeed faster than copying
files one at a time. I don't know the threshold, we should investigate with
different files to see it.
I think this should be included in C5.10 even if we don't use it for S3.
Encryption directories on HDFS will benefit from this too.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Vihang Karajgaonkar
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15536925#comment-15536925
]
Vihang Karajgaonkar commented on HIVE-14864:
Adding [~spena] since he has worked on this before.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Vihang Karajgaonkar
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15536922#comment-15536922
]
Vihang Karajgaonkar commented on HIVE-14864:
I looked into distcp documentation and I think you are right. It creates a list
of files to be copied if src is a directory and this list is divided among a
bunch of CopyMappers to do the actual copy. Do you have any ideas as to how
should we evaluate the threshold of number of files when distcp is beneficial?
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Vihang Karajgaonkar
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15536870#comment-15536870
]
Vihang Karajgaonkar commented on HIVE-14864:
Unfortunately, the documentation of ContentSummary.getLength() says ... returns
the length :)
This is the implementation of getContentSummary() in FileSystem.java which
suggests getLength is actually returning the sum of lengths of all the files
within that directory.
{noformat}
public ContentSummary getContentSummary(Path f) throws IOException {
FileStatus status = getFileStatus(f);
if (status.isFile()) {
// f is a file
return new ContentSummary(status.getLen(), 1, 0);
}
// f is a directory
long[] summary = {0, 0, 1};
for(FileStatus s : listStatus(f)) {
ContentSummary c = s.isDirectory() ? getContentSummary(s.getPath()) :
new ContentSummary(s.getLen(), 1, 0);
summary[0] += c.getLength();
summary[1] += c.getFileCount();
summary[2] += c.getDirectoryCount();
}
return new ContentSummary(summary[0], summary[1], summary[2]);
}
{noformat}
These are the revelant constructors for ContentSummary.
{noformat}
/** Constructor */
public ContentSummary(long length, long fileCount, long directoryCount) {
this(length, fileCount, directoryCount, -1L, length, -1L);
}
public ContentSummary(
long length, long fileCount, long directoryCount, long quota,
long spaceConsumed, long spaceQuota) {
this.length = length;
this.fileCount = fileCount;
this.directoryCount = directoryCount;
this.quota = quota;
this.spaceConsumed = spaceConsumed;
this.spaceQuota = spaceQuota;
}
{noformat}
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Vihang Karajgaonkar
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
[jira] [Commented] (HIVE-14864) Distcp is not called from MoveTask when src is a directory
[
https://issues.apache.org/jira/browse/HIVE-14864?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15536842#comment-15536842
]
Mohit Sabharwal commented on HIVE-14864:
Does srcFS.getContentSummary(src).getLength() return total number of files in
the directory ?
I think this whole condition needs to be re-thought. Because AFAIK, DistCp
speeds up copies when multiple
files are involved. There is no advantage in DistCp'ing a single file, no
matter how big that file. Which means
HIVE_EXEC_COPYFILE_MAXSIZE does not make sense even for a file.
IOW, we need to look into DistCp'ing directories only, and possibly ones which
contain more than a threshold
number of files.
> Distcp is not called from MoveTask when src is a directory
> --
>
> Key: HIVE-14864
> URL: https://issues.apache.org/jira/browse/HIVE-14864
> Project: Hive
> Issue Type: Bug
>Reporter: Vihang Karajgaonkar
>Assignee: Vihang Karajgaonkar
>
> In FileUtils.java the following code does not get executed even when src
> directory size is greater than HIVE_EXEC_COPYFILE_MAXSIZE because
> srcFS.getFileStatus(src).getLen() returns 0 when src is a directory. We
> should use srcFS.getContentSummary(src).getLength() instead.
> {noformat}
> /* Run distcp if source file/dir is too big */
> if (srcFS.getUri().getScheme().equals("hdfs") &&
> srcFS.getFileStatus(src).getLen() >
> conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE)) {
> LOG.info("Source is " + srcFS.getFileStatus(src).getLen() + " bytes.
> (MAX: " + conf.getLongVar(HiveConf.ConfVars.HIVE_EXEC_COPYFILE_MAXSIZE) +
> ")");
> LOG.info("Launch distributed copy (distcp) job.");
> HiveConfUtil.updateJobCredentialProviders(conf);
> copied = shims.runDistCp(src, dst, conf);
> if (copied && deleteSource) {
> srcFS.delete(src, true);
> }
> }
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
