[jira] [Created] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
Xu Cang created YARN-10516: -- Summary: In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception Key: YARN-10516 URL: https://issues.apache.org/jira/browse/YARN-10516 Project: Hadoop YARN Issue Type: Improvement Components: client Reporter: Xu Cang We have observed one issue from YARN client around this piece of code: [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] While {code:java} services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, defaultAddr, defaultPort)) .toString()); {code} Is being called, "yarnConf.getSocketAddr" will throw runtime exception, more specifically, UnknownHostException from here: [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] while one of the RM host was having networking issue that IP cannot be resolved. This runtime exception then floats all the way into our application and cause MR job submission failed. In my opinion, since we have HA here, multiple RMs are still alive and available. We should catch this exception in getTokenService() and handle it properly. Would like to hear your opinion on this, if agreed, I will provide a patch on this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Description: We have observed one issue from YARN client around this piece of code: [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] While {code:java} services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, defaultAddr, defaultPort)) .toString()); {code} is being called, "yarnConf.getSocketAddr" will throw runtime exception, more specifically, UnknownHostException from here: [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] while one of the RM host was having networking issue that IP cannot be resolved. This runtime exception then floats all the way up to our application and causes MR job submission failed. In my opinion, since we have HA here, multiple RMs are still alive and available. We should catch this exception in getTokenService() and handle it properly, instead of failing the whole action. Would like to hear your opinion on this, if agreed, I will provide a patch on this. Thank you. was: We have observed one issue from YARN client around this piece of code: [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] While {code:java} services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, defaultAddr, defaultPort)) .toString()); {code} Is being called, "yarnConf.getSocketAddr" will throw runtime exception, more specifically, UnknownHostException from here: [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] while one of the RM host was having networking issue that IP cannot be resolved. This runtime exception then floats all the way into our application and cause MR job submission failed. In my opinion, since we have HA here, multiple RMs are still alive and available. We should catch this exception in getTokenService() and handle it properly. Would like to hear your opinion on this, if agreed, I will provide a patch on this. Thank you. > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, "yarnConf.getSocketAddr" will throw runtime exception, > more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > while one of the RM host was having networking issue that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Description: We have observed one issue from YARN client around this piece of code: [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] While {code:java} services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, defaultAddr, defaultPort)) .toString()); {code} is being called, buildTokenService() fails and will throw runtime exception, more specifically, UnknownHostException from here: [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] *while one of the RM host was having networking issue* that IP cannot be resolved. This runtime exception then floats all the way up to our application and causes MR job submission failed. In my opinion, since we have HA here, multiple RMs are still alive and available. We should catch this exception in getTokenService() and handle it properly, instead of failing the whole action. Would like to hear your opinion on this, if agreed, I will provide a patch on this. Thank you. was: We have observed one issue from YARN client around this piece of code: [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] While {code:java} services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, defaultAddr, defaultPort)) .toString()); {code} is being called, "yarnConf.getSocketAddr" will throw runtime exception, more specifically, UnknownHostException from here: [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] while one of the RM host was having networking issue that IP cannot be resolved. This runtime exception then floats all the way up to our application and causes MR job submission failed. In my opinion, since we have HA here, multiple RMs are still alive and available. We should catch this exception in getTokenService() and handle it properly, instead of failing the whole action. Would like to hear your opinion on this, if agreed, I will provide a patch on this. Thank you. > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: (was: YARN-10516.001.patch) > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: YARN-10516.001.patch > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: YARN-10516.002.patch > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: YARN-10516.003.patch > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: YARN-10516.004.patch > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Updated] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Xu Cang updated YARN-10516: --- Attachment: YARN-10516.007.patch > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch, YARN-10516.007.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17247393#comment-17247393 ] Xu Cang commented on YARN-10516: [~epayne] [~hexiaoqiao] would you please review this Jira and patch? thanks! > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch, YARN-10516.007.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17247393#comment-17247393 ] Xu Cang edited comment on YARN-10516 at 12/10/20, 6:05 PM: --- [~epayne] [~hexiaoqiao] [~Jim_Brennan] would you please review this Jira and patch? thanks! was (Author: xucang): [~epayne] [~hexiaoqiao] would you please review this Jira and patch? thanks! > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch, YARN-10516.007.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17247393#comment-17247393 ] Xu Cang edited comment on YARN-10516 at 12/10/20, 11:50 PM: [~shahrs87] [~epayne] [~hexiaoqiao] [~Jim_Brennan] would you please review this Jira and patch? thanks! was (Author: xucang): [~epayne] [~hexiaoqiao] [~Jim_Brennan] would you please review this Jira and patch? thanks! > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch, YARN-10516.007.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org
[jira] [Commented] (YARN-10516) In HA mode, when one Resource Manager has networking issue, getTokenService() should not throw runtime exception
[ https://issues.apache.org/jira/browse/YARN-10516?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17253675#comment-17253675 ] Xu Cang commented on YARN-10516: [~epayne] [~hexiaoqiao] [~Jim_Brennan] Hi, would love to get some review on this, thank you > In HA mode, when one Resource Manager has networking issue, getTokenService() > should not throw runtime exception > > > Key: YARN-10516 > URL: https://issues.apache.org/jira/browse/YARN-10516 > Project: Hadoop YARN > Issue Type: Improvement > Components: client >Reporter: Xu Cang >Priority: Minor > Attachments: YARN-10516.001.patch, YARN-10516.002.patch, > YARN-10516.003.patch, YARN-10516.004.patch, YARN-10516.007.patch > > > We have observed one issue from YARN client around this piece of code: > [https://github.com/apache/hadoop/blob/trunk/hadoop-yarn-project/hadoop-yarn/hadoop-yarn-common/src/main/java/org/apache/hadoop/yarn/client/ClientRMProxy.java#L145] > > While > {code:java} > services.add(SecurityUtil.buildTokenService( yarnConf.getSocketAddr(address, > defaultAddr, defaultPort)) .toString()); > > {code} > is being called, buildTokenService() fails and will throw runtime > exception, more specifically, UnknownHostException from here: > [https://github.com/apache/hadoop/blob/trunk/hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/security/SecurityUtil.java#L466] > *while one of the RM host was having networking issue* that IP cannot be > resolved. > This runtime exception then floats all the way up to our application and > causes MR job submission failed. > In my opinion, since we have HA here, multiple RMs are still alive and > available. We should catch this exception in getTokenService() and handle it > properly, instead of failing the whole action. > > > Would like to hear your opinion on this, if agreed, I will provide a patch on > this. Thank you. -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org