Member since
10-09-2015
62
Posts
48
Kudos Received
14
Solutions
My Accepted Solutions
Title | Views | Posted |
---|---|---|
3231 | 09-07-2017 07:15 PM | |
3175 | 06-26-2017 08:24 PM | |
2370 | 02-03-2017 08:21 PM | |
2845 | 01-28-2017 12:32 AM | |
2188 | 01-12-2017 07:39 PM |
09-07-2017
07:15 PM
3 Kudos
@Chingiz Akmatov spark-env.sh is ambari managed file and expose on UI. So you can make changes to it from Amabri UI. To do so Go to Spark2->Configs->Advanced spark2-env section->content property and add the env variable
... View more
06-26-2017
08:24 PM
@David Pocivalnik every time recommendation API call is made, new entry for logs is made at path "/var/run/ambari-server/stack-recommendations/" on ambari-server host. there will be several and if you are not certain about the last entry then please move all historical existing entries to different location and then again go through add service wizard to produce a repro. Entry will have stackadvisor.err and stackadvisor.out files. Please refer them for further debugging the issue. If it's not clear from logs, please compress the directory and post it on this thread, So that we can get more information for debugging purpose.
... View more
02-27-2017
05:35 PM
2 Kudos
@Sachin Ambardekar From HDFS perspective, in some rare circumstances it was noticed that secondary (or standby) namenode fails to consume edit log. This results in more complicated situations if active namenode is restarted meanwhile (unconsumed edit logs will have to be ignored). The simpler solution to handle such scenario more gracefully is to always make sure that fsimage is updated before stopping namenode. So as precautionary measure work was done in Ambari to check and warn user if user tries to stop NameNode that has a checkpoint older than 12 hours. [1] HDFS-3.0.0.0 has implemented this check natively and going forward Ambari might skip this warning. [2] Following Jira's and their description are used as references for this answer: [1] https://issues.apache.org/jira/browse/AMBARI-12951 [2] https://issues.apache.org/jira/browse/HDFS-6353
... View more
02-10-2017
11:38 AM
1 Kudo
@will chen you can add spark johistoryserver back by using ambari-web UI: Go to the host details page. This can be done by clicking on the hosts tab and clicking on the host where you want to install spark jobhistoryserver Click on Add button. It should show list of components that can be added to the host. Click on spark jobhistoryserver. please see the attached image add-component.png If for any reason you are not able to add it from UI or you want to use ambari-server REST APIs to do so then article at link has a section "Step 5 - Create host components" which can be used to add host components. another question: I cannot find SPARK_THRIFTSERVER in /var/lib/ambari-server/resources/common-services/SPARK/1.2.0.2.2/metainfo.xml. Why does it exists in SPARK service?
Thriftserver support for spark was added first in spark 1.4.1 definition in ambari (link). It will be defined at /var/lib/ambari-server/resources/common-services/SPARK/1.4.1/metainfo.xml.
... View more
02-07-2017
11:27 PM
@Meng Meng Glad to know that. Please accept the above answer when you get chance, so it can be noted as correct answer to this issue.
... View more
02-03-2017
08:21 PM
2 Kudos
Hi @Meng Meng Ambari UI si completely driven from Ambari REST API API call to create a new YARN config group and add a host from default group to it: curl 'http://c6401.ambari.apache.org:8080/api/v1/clusters/c1/config_groups' -u admin:admin -H "X-Requested-By: ambari" -i -X POST --data '[{"ConfigGroup":{"group_name":"custom_group","tag":"YARN","description":"This is a custom group","desired_configs":[],"hosts":[{"host_name":"c6401.ambari.apache.org"}]}}]'
But if a host does not belong to Default config group and belongs to any other custom group and is desired to be moved to another custom config group then first it needs to be removed from prior custom config group and then be added to the desired new custom config group. The same API (use PUT instead of POST method) as mentioned above can be used to first remove it from the previous config group curl 'http://c6401.ambari.apache.org:8080/api/v1/clusters/c1/config_groups/2' -u admin:admin -H "X-Requested-By: ambari" -i -X PUT --data '{"ConfigGroup":{"group_name":"custom_group","description":"This is a custom group","tag":"YARN","hosts":[],"desired_configs":[]}}' --compressed Note: In above mentioned APIs, you need to change ambari-server hostname, cluster name, group_name, description and tag (service name) as applicable in your environment
... View more
01-31-2017
07:19 PM
@Sami Ahmad can you let us know the value of the property when yarn RM HA was done. This can be achieved by using the API is described in my last comment http://localhost:8080/api/v1/clusters/c1/configurations/service_config_versions?service_name=YARN&service_config_version_note=This%20configuration%20is%20created%20by%20Enable%20ResourceManager%20HA%20wizard I am trying to understand if that yarn property was set to true when RAM HA was done for the very first time and then later got reverted to false OR value of that property was false even when RM HA was completed We can know that by looking into the property value by using above API as that service config version is created when YARN RM HA is completed. Also let us know ambari version. It allows us to look into the correct version of ambari code and verify if it's a bug specific to the ambari version that you are using or not.
... View more
01-28-2017
12:32 AM
@Sami Ahmad 1. can you please let us know the version of ambari that you are using ? 2. Also can you check the current value for the same property (yarn.resourcemanager.ha.enabled) on YARN service config page in Ambari ? 3. Can you view the config version that enabled RM HA and check the same property's value ? you can do this from UI and API. UI: enable-rm-ha.png API: http://localhost:8080/api/v1/clusters/c1/configurations/service_config_versions?service_name=YARN&service_config_version_note=This%20configuration%20is%20created%20by%20Enable%20ResourceManager%20HA%20wizard Once you get the response from API, do a lookup for the property value
... View more
01-17-2017
10:56 PM
1 Kudo
@Sami Ahmad ambari-web uses below pointed code to determine enabling of next button: https://github.com/apache/ambari/blob/release-2.4.0/ambari-web/app/controllers/main/admin/highAvailability/nameNode/step4_controller.js#L45-L55 Following API call will help to debug further the reason behind this issue: http://c6401.ambari.apache.org:8080/api/v1/clusters/c1/hosts/{NAMENODE_HOSTNAME}/host_components/NAMENODE?fields=metrics/dfs/namenode/Safemode,metrics/dfs/namenode/JournalTransactionInfo Value of Safemode and JournalTransactionInfo should not be empty. value of LastAppliedOrWrittenTxId - value of MostRecentCheckpointTxId should be less than or equal to 1
... View more
01-12-2017
07:39 PM
1 Kudo
@Shashant Panwar you will also need to clean alerts and service config tables: Some of them are as followed: serviceconfig serviceconfigmapping alert_definition alert_current alert_group alert_history I will rather encourage to resolve the issue of showing Mahout in the UI and then delete service from UI (if using ambari-2.4.0 or above) or from Ambari API Can you let us know what version of ambari are you using ? Also what is the content of clusterservices and servicedesiredstate tables specifically if Mahout service entry is present or not?
... View more