-
Notifications
You must be signed in to change notification settings - Fork 174
Updated hadoop version from 2.6.0 to 3.1.0 #607
base: master
Are you sure you want to change the base?
Conversation
@@ -119,8 +119,8 @@ | |||
@Deprecated | |||
public static final String KEY_TAB_FILE = StramUserLogin.DT_AUTH_PREFIX + "store.keytab"; | |||
public static final String TOKEN_ANTICIPATORY_REFRESH_FACTOR = StramUserLogin.DT_AUTH_PREFIX + "token.refresh.factor"; | |||
public static final long DELEGATION_TOKEN_MAX_LIFETIME_DEFAULT = 7 * 24 * 60 * 60 * 1000; | |||
public static final long DELEGATION_TOKEN_RENEW_INTERVAL_DEFAULT = 24 * 60 * 60 * 1000; | |||
public static final long RM_DELEGATION_TOKEN_MAX_LIFETIME_DEFAULT = 7 * 24 * 60 * 60 * 1000; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Suggest you submit this change in a separate PR as it is not related to the Hadoop 3.1 change.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I seem to remember Intellij complaining about those two lines after I made the changes to the constants used by Hadoop 3.1. Perhaps I misunderstood/misinterpreted the relationship?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is not necessary to rename Apex constants.
* exception java.io.IOException is never thrown in body of corresponding try statement | ||
* The remedy for this appears to be to use a more specific exception as the current try/catch block could | ||
* not spit out IOException. For now, just using Exception to try and get this to build, will revisit later. */ | ||
} catch (Exception ex) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Have you tried removing the catch clause completely and let your IDE insert the catch clause(s) for the exact specific types? Exception is more general than IOException but includes RuntimeException, is that why the compiler likes it?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ah, that is a good idea...should have thought of it....I'll give that a shot...as it relates to why Exception made the compiler happy...I suspect because Exception covers pretty much everything, it is a horrible choice, but not, from a compiler's standpoint, incorrect.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Remove try/catch as getPos()
does not throw IOException
now.
* been deprecated as far back as I can find Hadoop versions (2.4.1 at least), HADOOP_CALLER_CONTEXT_ENABLED_DEFAULT and | ||
* HADOOP_CALLER_CONTEXT_ENABLED_KEY are the only two constants in 3.1.0 that seem even remotely related... | ||
* */ | ||
boolean sslEnabled = conf.getBoolean(CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_KEY, CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_DEFAULT); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
My comments on this:
- instead of having a tentative fix like this why can't we have a permanent fix? See below
- Do we know what string constants these identifiers CommonConfigurationKeysPublic.HADOOP_SSL_ENABLED_KEY and CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_KEY stand for?
- Same thing for CommonConfigurationKeysPublic.HADOOP_SSL_ENABLED_DEFAULT and CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_DEFAULT
- if they are the same (which I doubt) we haven't introduced any incompatibility as far as external user impact is concerned.
- if they are different we can just break the compatibility (not desirable) by using these new property names
- or (desirable) honor the old property value by defining it here and checking for it before checking the new value.\
- the comment is also repeated elsewhere. Ideally it should only exist in one place.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
You make a good point...wil look into the constants more...the weird thing is that in the Javadocs for Hadoop, those two constants have no information about them as far back as I can see...other than being marked as deprecated...will do some more digging. From the APEX side of the house, what are we doing with this function other than the obvious...checking to see if SSL is enabled.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't know how reliable this page is...but the string/boolean values for the constants seem to make sense...I will hunt down which constants, if any, replaced them...
public static final String HADOOP_SSL_ENABLED_KEY = "hadoop.ssl.enabled"
public static final boolean HADOOP_SSL_ENABLED_DEFAULT = false
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It might be worthwhile asking on the users@ list if anyone is using Apex in SSL mode. If no-one is using SSL we don't need to worry about breaking compatibility. But we should still confirm that HADOOP_CALLER_CONTEXT_ENABLED_KEY is the right property to check to verify SSL is enabled in the user's Hadoop cluster.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Oh, I think our comments crossed. And what's the value of HADOOP_CALLER_CONTEXT_ENABLED_KEY
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is not sufficient to ask on "user@apex" who uses SSL. It is necessary to vote on "dev@apex" prior to dropping SSL support and I will be the one to -1 it.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@vrozov who is recommending dropping SSL support? I didn't suggest that if that's the impression you got. My question was in connection with creating incompatibility which might be okay if no-one is currently using SSL
Both the Jenkins and Travis builds/tests have failed. Could you check the failures and see if they are related to these changes? |
Yeah, I was just looking into that. I reverted all my changes, so I have a vanilla Master branch...the build fails there too with the tests: <failure message="Version number pattern does not match: Unknown" type="java.lang.AssertionError">java.lang.AssertionError: Version number pattern does not match: Unknown
at com.datatorrent.stram.util.VersionInfoTest.testMavenProperties(VersionInfoTest.java:64)
</failure>
|
If this is the same failure you see in your PR CI tests then that's good news. Not sure why we are getting "Unknown" from the VersionInfo object but it might be worthwhile debugging and fixing that in a separate PR. |
@abossert Please open JIRA and follow "Opening Pull Requests" guidelines (see http://apex.apache.org/contributing.html) |
@@ -14,8 +14,7 @@ | |||
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY | |||
* KIND, either express or implied. See the License for the | |||
* specific language governing permissions and limitations | |||
* under the License. | |||
*/ | |||
* under the License.*/ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
undo this formatting change.
@@ -119,8 +119,8 @@ | |||
@Deprecated | |||
public static final String KEY_TAB_FILE = StramUserLogin.DT_AUTH_PREFIX + "store.keytab"; | |||
public static final String TOKEN_ANTICIPATORY_REFRESH_FACTOR = StramUserLogin.DT_AUTH_PREFIX + "token.refresh.factor"; | |||
public static final long DELEGATION_TOKEN_MAX_LIFETIME_DEFAULT = 7 * 24 * 60 * 60 * 1000; | |||
public static final long DELEGATION_TOKEN_RENEW_INTERVAL_DEFAULT = 24 * 60 * 60 * 1000; | |||
public static final long RM_DELEGATION_TOKEN_MAX_LIFETIME_DEFAULT = 7 * 24 * 60 * 60 * 1000; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is not necessary to rename Apex constants.
* been deprecated as far back as I can find Hadoop versions (2.4.1 at least), HADOOP_CALLER_CONTEXT_ENABLED_DEFAULT and | ||
* HADOOP_CALLER_CONTEXT_ENABLED_KEY are the only two constants in 3.1.0 that seem even remotely related... | ||
* */ | ||
boolean sslEnabled = conf.getBoolean(CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_KEY, CommonConfigurationKeysPublic.HADOOP_CALLER_CONTEXT_ENABLED_DEFAULT); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is not sufficient to ask on "user@apex" who uses SSL. It is necessary to vote on "dev@apex" prior to dropping SSL support and I will be the one to -1 it.
* exception java.io.IOException is never thrown in body of corresponding try statement | ||
* The remedy for this appears to be to use a more specific exception as the current try/catch block could | ||
* not spit out IOException. For now, just using Exception to try and get this to build, will revisit later. */ | ||
} catch (Exception ex) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Remove try/catch as getPos()
does not throw IOException
now.
@abossert just checking. Hopefully you are not stuck on anything and making progress on this PR. |
Sorry, have been bed ridden for a day with a nasty flu and swamped with work...I am still not entirely sure where to look for the version error that needs to be fixed (prevents initial compile and test from running prior to any Hadoop version change)...digging into it today, but would appreciate any thoughts you might have?
…Sent from my iPhone
On Jan 23, 2019, at 23:53, sanjaypujare ***@***.***> wrote:
@abossert just checking. Hopefully you are not stuck on anything and making progress on this PR.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
@tweise added that test (com.datatorrent.stram.util.VersionInfoTest.testMavenProperties). Any ideas? |
OK, so in com.datatorrent.stram.util.VersionInfo on line 61, it is not
getting populated and is not producing any error either, but rather staying
with the default values of Unknown:
Enumeration<URL> resources =
classInJar.getClassLoader().getResources("META-INF/maven/" + groupId +
"/" + artifactId + "/pom.properties");
…On Thu, Jan 24, 2019 at 12:11 PM sanjaypujare ***@***.***> wrote:
@tweise <https://github.com/tweise> added that test
(com.datatorrent.stram.util.VersionInfoTest.testMavenProperties). Any ideas?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#607 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AYuy1DZUdFYG0uWU61iTlmSwXZOjWbdyks5vGek5gaJpZM4aCUt->
.
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
@abossert looks like a different error than what is shown in the CI. How are you running the tests? Try |
I am at a loss. Tried completely restarting and all seems fine now. Got
the initial build to work pre-upgrade to hadoop 3.1.0:
[DEBUG] Using catalog /Users/mbossert/.m2/archetype-catalog.xml
[DEBUG] Reading the catalog /Users/mbossert/.m2/archetype-catalog.xml
[INFO]
------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO]
[INFO] Apache Apex ........................................ SUCCESS [
2.744 s]
[INFO] Apache Apex API .................................... SUCCESS [
8.776 s]
[INFO] Apache Apex Common Library ......................... SUCCESS [
6.235 s]
[INFO] Apache Apex Buffer Server .......................... SUCCESS [
4.226 s]
[INFO] Apache Apex Stream Processing Engine ............... SUCCESS [08:27
min]
[INFO] Apache Apex Application Maven Archetype ............ SUCCESS [
1.977 s]
[INFO] Apache Apex App Configuration Maven Archetype ...... SUCCESS [
0.613 s]
[INFO]
------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO]
------------------------------------------------------------------------
[INFO] Total time: 08:52 min
[INFO] Finished at: 2019-01-24T16:16:35-05:00
[INFO] Final Memory: 100M/1516M
[INFO]
------------------------------------------------------------------------
Process finished with exit code 0
…On Thu, Jan 24, 2019 at 2:57 PM Thomas Weise ***@***.***> wrote:
@abossert <https://github.com/abossert> looks like a different error than
what is shown in the CI. How are you running the tests?
Try mvn clean install if that's not what you are doing already.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#607 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AYuy1EmSUgjp__Qycyq77RnAEEB3SHDuks5vGhAegaJpZM4aCUt->
.
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
Another question...I have been able to validate that all of the constant
values that were causing build errors have been replaced with the
appropriate replacements from Hadoop 2.6.1 -> 3.1.0, however the two
following ones just show up as deprecated with no clue about a
replacement. I found a reference to them in Hadoop 2.5 source code...but
not sure where that would leave us...can anyone chime in on background for
how these variables are used in Apex?
@deprecated
/** Only used by HttpServer. */
public static final String HADOOP_SSL_ENABLED_KEY = "hadoop.ssl.enabled";
@deprecated
/** Only used by HttpServer. */
public static final boolean HADOOP_SSL_ENABLED_DEFAULT = false;
…On Thu, Jan 24, 2019 at 4:18 PM Aaron Bossert ***@***.***> wrote:
I am at a loss. Tried completely restarting and all seems fine now. Got
the initial build to work pre-upgrade to hadoop 3.1.0:
[DEBUG] Using catalog /Users/mbossert/.m2/archetype-catalog.xml
[DEBUG] Reading the catalog /Users/mbossert/.m2/archetype-catalog.xml
[INFO]
------------------------------------------------------------------------
[INFO] Reactor Summary:
[INFO]
[INFO] Apache Apex ........................................ SUCCESS [
2.744 s]
[INFO] Apache Apex API .................................... SUCCESS [
8.776 s]
[INFO] Apache Apex Common Library ......................... SUCCESS [
6.235 s]
[INFO] Apache Apex Buffer Server .......................... SUCCESS [
4.226 s]
[INFO] Apache Apex Stream Processing Engine ............... SUCCESS [08:27
min]
[INFO] Apache Apex Application Maven Archetype ............ SUCCESS [
1.977 s]
[INFO] Apache Apex App Configuration Maven Archetype ...... SUCCESS [
0.613 s]
[INFO]
------------------------------------------------------------------------
[INFO] BUILD SUCCESS
[INFO]
------------------------------------------------------------------------
[INFO] Total time: 08:52 min
[INFO] Finished at: 2019-01-24T16:16:35-05:00
[INFO] Final Memory: 100M/1516M
[INFO]
------------------------------------------------------------------------
Process finished with exit code 0
On Thu, Jan 24, 2019 at 2:57 PM Thomas Weise ***@***.***>
wrote:
> @abossert <https://github.com/abossert> looks like a different error
> than what is shown in the CI. How are you running the tests?
>
> Try mvn clean install if that's not what you are doing already.
>
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub
> <#607 (comment)>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/AYuy1EmSUgjp__Qycyq77RnAEEB3SHDuks5vGhAegaJpZM4aCUt->
> .
>
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
nevermind...I found the replacement...but not sure how you guys would like
to tackle it...any thoughts?
ParameterValueNotes
dfs.http.policy HTTP_ONLY or HTTPS_ONLY or HTTP_AND_HTTPS HTTPS_ONLY turns
off http access. This option takes precedence over the deprecated
configuration dfs.https.enable and hadoop.ssl.enabled. If using SASL to
authenticate data transfer protocol instead of running DataNode as root and
using privileged ports, then this property must be set to HTTPS_ONLY to
guarantee authentication of HTTP servers. (See dfs.data.transfer.protection
.)
…On Thu, Jan 24, 2019 at 5:21 PM Aaron Bossert ***@***.***> wrote:
Another question...I have been able to validate that all of the constant
values that were causing build errors have been replaced with the
appropriate replacements from Hadoop 2.6.1 -> 3.1.0, however the two
following ones just show up as deprecated with no clue about a
replacement. I found a reference to them in Hadoop 2.5 source code...but
not sure where that would leave us...can anyone chime in on background for
how these variables are used in Apex?
@deprecated
/** Only used by HttpServer. */
public static final String HADOOP_SSL_ENABLED_KEY = "hadoop.ssl.enabled";
@deprecated
/** Only used by HttpServer. */
public static final boolean HADOOP_SSL_ENABLED_DEFAULT = false;
On Thu, Jan 24, 2019 at 4:18 PM Aaron Bossert ***@***.***>
wrote:
> I am at a loss. Tried completely restarting and all seems fine now. Got
> the initial build to work pre-upgrade to hadoop 3.1.0:
>
> [DEBUG] Using catalog /Users/mbossert/.m2/archetype-catalog.xml
> [DEBUG] Reading the catalog /Users/mbossert/.m2/archetype-catalog.xml
> [INFO]
> ------------------------------------------------------------------------
> [INFO] Reactor Summary:
> [INFO]
> [INFO] Apache Apex ........................................ SUCCESS [
> 2.744 s]
> [INFO] Apache Apex API .................................... SUCCESS [
> 8.776 s]
> [INFO] Apache Apex Common Library ......................... SUCCESS [
> 6.235 s]
> [INFO] Apache Apex Buffer Server .......................... SUCCESS [
> 4.226 s]
> [INFO] Apache Apex Stream Processing Engine ............... SUCCESS
> [08:27 min]
> [INFO] Apache Apex Application Maven Archetype ............ SUCCESS [
> 1.977 s]
> [INFO] Apache Apex App Configuration Maven Archetype ...... SUCCESS [
> 0.613 s]
> [INFO]
> ------------------------------------------------------------------------
> [INFO] BUILD SUCCESS
> [INFO]
> ------------------------------------------------------------------------
> [INFO] Total time: 08:52 min
> [INFO] Finished at: 2019-01-24T16:16:35-05:00
> [INFO] Final Memory: 100M/1516M
> [INFO]
> ------------------------------------------------------------------------
>
> Process finished with exit code 0
>
> On Thu, Jan 24, 2019 at 2:57 PM Thomas Weise ***@***.***>
> wrote:
>
>> @abossert <https://github.com/abossert> looks like a different error
>> than what is shown in the CI. How are you running the tests?
>>
>> Try mvn clean install if that's not what you are doing already.
>>
>> —
>> You are receiving this because you were mentioned.
>> Reply to this email directly, view it on GitHub
>> <#607 (comment)>,
>> or mute the thread
>> <https://github.com/notifications/unsubscribe-auth/AYuy1EmSUgjp__Qycyq77RnAEEB3SHDuks5vGhAegaJpZM4aCUt->
>> .
>>
>
>
> --
>
> M. Aaron Bossert
> (202) 594-1703
> Punch Cyber Analytics Group
>
>
>
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
My preference: just make the change to honor dfs.http.policy only and not worry about dfs.https.enable and hadoop.ssl.enabled . We haven't heard about any existing Apex users using SSL. |
Given that the new one is ternary (https, http, and either) and the old one is Boolean, how would suggest to handle it? True == HTTPS, false == HTTP, and then err towards true in case of either?
…Sent from my iPhone
On Jan 24, 2019, at 19:16, sanjaypujare ***@***.***> wrote:
My preference: just make the change to honor dfs.http.policy only and not worry about dfs.https.enable and hadoop.ssl.enabled . We haven't heard about any existing Apex users using SSL.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
Good question. As per https://hadoop.apache.org/docs/r3.1.0/hadoop-project-dist/hadoop-hdfs/hdfs-default.xml, |
OK, so I am past this part...on to the next error:
Running com.datatorrent.stram.StramMiniClusterTest
Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.003 sec
<<< FAILURE! - in com.datatorrent.stram.StramMiniClusterTest
com.datatorrent.stram.StramMiniClusterTest Time elapsed: 0.003 sec <<<
ERROR!
java.lang.NoClassDefFoundError: org/apache/hadoop/net/ServerSocketUtil
at
com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
Caused by: java.lang.ClassNotFoundException:
org.apache.hadoop.net.ServerSocketUtil
at
com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
Does this make sense? When I look at org.apache.hadoop.net in the Javadoc,
I can't find it even in the 2.6.0 version, which compiled just fine...
…On Thu, Jan 24, 2019 at 11:07 PM sanjaypujare ***@***.***> wrote:
Good question. As per
https://hadoop.apache.org/docs/r3.1.0/hadoop-project-dist/hadoop-hdfs/hdfs-default.xml,
HTTP_AND_HTTPS means service is provided both on http and https. We use
this setting for both outbound and inbound (apex web-service) and enhancing
our inbound to support HTTP_AND_HTTPS will be some work which is not
really needed at this time. I am okay to treat HTTP_AND_HTTPS as
HTTPS_ONLY.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#607 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/AYuy1FQexz08SfhioRtZ4o95VbIHluAKks5vGoMNgaJpZM4aCUt->
.
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
Sorry, by "it", I meant the ServerSocketUtil class
…On Fri, Jan 25, 2019 at 1:58 AM Aaron Bossert ***@***.***> wrote:
OK, so I am past this part...on to the next error:
Running com.datatorrent.stram.StramMiniClusterTest
Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.003 sec
<<< FAILURE! - in com.datatorrent.stram.StramMiniClusterTest
com.datatorrent.stram.StramMiniClusterTest Time elapsed: 0.003 sec <<<
ERROR!
java.lang.NoClassDefFoundError: org/apache/hadoop/net/ServerSocketUtil
at
com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
Caused by: java.lang.ClassNotFoundException:
org.apache.hadoop.net.ServerSocketUtil
at
com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
Does this make sense? When I look at org.apache.hadoop.net in the
Javadoc, I can't find it even in the 2.6.0 version, which compiled just
fine...
On Thu, Jan 24, 2019 at 11:07 PM sanjaypujare ***@***.***>
wrote:
> Good question. As per
> https://hadoop.apache.org/docs/r3.1.0/hadoop-project-dist/hadoop-hdfs/hdfs-default.xml,
> HTTP_AND_HTTPS means service is provided both on http and https. We use
> this setting for both outbound and inbound (apex web-service) and enhancing
> our inbound to support HTTP_AND_HTTPS will be some work which is not
> really needed at this time. I am okay to treat HTTP_AND_HTTPS as
> HTTPS_ONLY.
>
> —
> You are receiving this because you were mentioned.
> Reply to this email directly, view it on GitHub
> <#607 (comment)>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/AYuy1FQexz08SfhioRtZ4o95VbIHluAKks5vGoMNgaJpZM4aCUt->
> .
>
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
Running into another build/test issue...The testing has been running for
hours and seems to be stuck in an endless loop of the following...Does this
make sense to anyone?
2019-01-25 20:35:44,796 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:45,803 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:46,811 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:47,819 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:48,830 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:49,837 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:50,846 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:51,852 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:52,860 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:53,864 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:54,875 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:55,883 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:56,890 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:57,900 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:58,907 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:35:59,917 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:00,925 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:01,930 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:02,933 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:03,942 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:04,952 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:05,960 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:06,966 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:07,968 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:08,976 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:09,985 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:10,994 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:12,003 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:13,013 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:14,020 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:15,030 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:16,041 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:17,051 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:18,061 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:19,070 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:20,076 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:21,080 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:22,084 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:23,093 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:24,100 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:25,103 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:26,111 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:27,119 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:28,121 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:29,127 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:30,136 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:31,145 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:32,149 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:33,158 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:34,167 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:35,177 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:36,186 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:37,188 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:38,190 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:39,195 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:40,197 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:41,201 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:42,203 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:43,207 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:44,211 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:45,216 [master] WARN stram.StreamingContainerManager
updateRecoveryCheckpoints - Marking operator
PTOperator[id=2,name=o1,state=ACTIVE] blocked committed window
ffffffffffffffff, recovery window ffffffffffffffff, current time
1548466605215, last window id change time 1548466544796, window processing
timeout millis 60000
2019-01-25 20:36:45,216 [master] INFO stram.StreamingContainerManager
updateCheckpoints - Blocked operator PTOperator[id=2,name=o1,state=ACTIVE]
container PTContainer[id=2(container-166),state=ACTIVE] time 60419ms
2019-01-25 20:36:45,216 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:46,217 [master] INFO engine.StreamingContainer
processHeartbeatResponse - Received shutdown request type ABORT
2019-01-25 20:36:46,217 [master] INFO stram.StramLocalCluster run -
Container container-166 restart.
2019-01-25 20:36:46,217 [master] INFO stram.StreamingContainerManager
scheduleContainerRestart - Initiating recovery for container-166@localhost
2019-01-25 20:36:46,217 [master] WARN stram.StreamingContainerManager
updateRecoveryCheckpoints - Marking operator
PTOperator[id=2,name=o1,state=ACTIVE] blocked committed window
ffffffffffffffff, recovery window ffffffffffffffff, current time
1548466606217, last window id change time 1548466544796, window processing
timeout millis 60000
2019-01-25 20:36:46,217 [master] INFO stram.StreamingContainerManager
scheduleContainerRestart - Affected operators
[PTOperator[id=2,name=o1,state=ACTIVE]]
2019-01-25 20:36:46,219 [master] INFO stram.StreamingContainerManager
assignContainer - Removing container agent container-166
2019-01-25 20:36:46,219 [container-167] INFO stram.StramLocalCluster run -
Started container container-167
2019-01-25 20:36:46,219 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:46,220 [container-167] INFO stram.StramLocalCluster log -
container-167 msg: [container-167] Entering heartbeat loop..
2019-01-25 20:36:47,161 [container-166] ERROR
stram.StreamingContainerManager processHeartbeat - Unknown container
container-166
2019-01-25 20:36:47,161 [container-166] INFO engine.StreamingContainer
processHeartbeatResponse - Received shutdown request type ABORT
2019-01-25 20:36:47,161 [container-166] INFO stram.StramLocalCluster log -
container-166 msg: [container-166] Exiting heartbeat loop..
2019-01-25 20:36:47,164 [container-166] INFO stram.StramLocalCluster run -
Container container-166 terminating.
2019-01-25 20:36:47,224 [master] WARN stram.StreamingContainerManager
calculateEndWindowStats - Some operators are behind for more than 1000
windows! Trimming the end window stats map
2019-01-25 20:36:47,225 [container-167] INFO engine.StreamingContainer
processHeartbeatResponse - Deploy request:
[OperatorDeployInfo[id=2,name=o1,type=GENERIC,checkpoint={ffffffffffffffff,
0,
0},inputs=[OperatorDeployInfo.InputDeployInfo[portName=inport1,streamId=TestTuples,sourceNodeId=1,sourcePortName=outport,locality=<null>,partitionMask=0,partitionKeys=<null>]],outputs=[]]]
2019-01-25 20:36:47,225 [container-167] INFO engine.StreamingContainer
deployNodes - Restoring operator 2 to checkpoint ffffffffffffffff
stateless=false.
…On Fri, Jan 25, 2019 at 2:01 AM Aaron Bossert ***@***.***> wrote:
Sorry, by "it", I meant the ServerSocketUtil class
On Fri, Jan 25, 2019 at 1:58 AM Aaron Bossert ***@***.***>
wrote:
> OK, so I am past this part...on to the next error:
>
> Running com.datatorrent.stram.StramMiniClusterTest
> Tests run: 1, Failures: 0, Errors: 1, Skipped: 0, Time elapsed: 0.003 sec
> <<< FAILURE! - in com.datatorrent.stram.StramMiniClusterTest
> com.datatorrent.stram.StramMiniClusterTest Time elapsed: 0.003 sec <<<
> ERROR!
> java.lang.NoClassDefFoundError: org/apache/hadoop/net/ServerSocketUtil
> at
> com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
> Caused by: java.lang.ClassNotFoundException:
> org.apache.hadoop.net.ServerSocketUtil
> at
> com.datatorrent.stram.StramMiniClusterTest.setup(StramMiniClusterTest.java:139)
>
> Does this make sense? When I look at org.apache.hadoop.net in the
> Javadoc, I can't find it even in the 2.6.0 version, which compiled just
> fine...
>
> On Thu, Jan 24, 2019 at 11:07 PM sanjaypujare ***@***.***>
> wrote:
>
>> Good question. As per
>> https://hadoop.apache.org/docs/r3.1.0/hadoop-project-dist/hadoop-hdfs/hdfs-default.xml,
>> HTTP_AND_HTTPS means service is provided both on http and https. We use
>> this setting for both outbound and inbound (apex web-service) and enhancing
>> our inbound to support HTTP_AND_HTTPS will be some work which is not
>> really needed at this time. I am okay to treat HTTP_AND_HTTPS as
>> HTTPS_ONLY.
>>
>> —
>> You are receiving this because you were mentioned.
>> Reply to this email directly, view it on GitHub
>> <#607 (comment)>,
>> or mute the thread
>> <https://github.com/notifications/unsubscribe-auth/AYuy1FQexz08SfhioRtZ4o95VbIHluAKks5vGoMNgaJpZM4aCUt->
>> .
>>
>
>
> --
>
> M. Aaron Bossert
> (202) 594-1703
> Punch Cyber Analytics Group
>
>
>
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
--
M. Aaron Bossert
(202) 594-1703
Punch Cyber Analytics Group
|
@abossert thanks for the contribution. Seeing no activity on the PR since Jan, what's the plan forward? |
Thomas,
It’s still on my radar...but have been sidetracked with a large project at work. I’m circling back around to this later this week.
Aaron
…Sent from my iPhone
On Apr 13, 2019, at 15:52, Thomas Weise ***@***.***> wrote:
@abossert thanks for the contribution. Seeing no activity on the PR since Jan, what's the plan forward?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
|
@abossert any update? If not, I'll close the PR, please re-open when there is an update. |
As discussed on the dev mail list, I have upgraded Hadoop from 2.6.0 to 3.1.0. The build is successful, but no other functionality or incorporation of new Hadoop 3.1 features yet...