Re: Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-31 Thread Alejandro Fernandez 


> On March 31, 2016, 1:13 a.m., Alejandro Fernandez wrote:
> > ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java,
> >  line 403
> > 
> >
> > This will acquire a read lock every time. If performance suffers, 
> > perhaps we can relax that constraint a bit since alerts have a bit more 
> > leeway
> 
> Jonathan Hurley wrote:
> I'm fine relaxing this a bit. In fact, I think we're negating the 
> benefits of the ConcurrentHashMap here.

+1 on removing the readlock


- Alejandro


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/#review126231
---


On March 29, 2016, 7:32 p.m., Jonathan Hurley wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45442/
> ---
> 
> (Updated March 29, 2016, 7:32 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
> Cole.
> 
> 
> Bugs: AMBARI-15620
> https://issues.apache.org/jira/browse/AMBARI-15620
> 
> 
> Repository: ambari
> 
> 
> Description
> ---
> 
> Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after 
> removing a host because they are always considered valid. 
> 
> STR
> - Deploy cluster 
> - Add/Remove nodes a few times 
> - Removed all aded nodes
> 
> {code}
>  There are 4 stale alerts from 4 host(s): 
> amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
> {code}
> 
> 
> Diffs
> -
> 
>   
> ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
>  8dc8e1e 
>   ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
> ebd29e3 
>   ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
> a1ebaba 
>   
> ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
>  6c68d0e 
>   
> ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
>  136a756 
> 
> Diff: https://reviews.apache.org/r/45442/diff/
> 
> 
> Testing
> ---
> 
> mvn clean test
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Re: Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-31 Thread Jonathan Hurley 


> On March 30, 2016, 9:13 p.m., Alejandro Fernandez wrote:
> > ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java,
> >  line 403
> > 
> >
> > This will acquire a read lock every time. If performance suffers, 
> > perhaps we can relax that constraint a bit since alerts have a bit more 
> > leeway

I'm fine relaxing this a bit. In fact, I think we're negating the benefits of 
the ConcurrentHashMap here.


- Jonathan


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/#review126231
---


On March 29, 2016, 3:32 p.m., Jonathan Hurley wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45442/
> ---
> 
> (Updated March 29, 2016, 3:32 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
> Cole.
> 
> 
> Bugs: AMBARI-15620
> https://issues.apache.org/jira/browse/AMBARI-15620
> 
> 
> Repository: ambari
> 
> 
> Description
> ---
> 
> Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after 
> removing a host because they are always considered valid. 
> 
> STR
> - Deploy cluster 
> - Add/Remove nodes a few times 
> - Removed all aded nodes
> 
> {code}
>  There are 4 stale alerts from 4 host(s): 
> amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
> {code}
> 
> 
> Diffs
> -
> 
>   
> ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
>  8dc8e1e 
>   ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
> ebd29e3 
>   ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
> a1ebaba 
>   
> ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
>  6c68d0e 
>   
> ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
>  136a756 
> 
> Diff: https://reviews.apache.org/r/45442/diff/
> 
> 
> Testing
> ---
> 
> mvn clean test
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Re: Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-30 Thread Alejandro Fernandez 

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/#review126231
---




ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
 (line 403)


This will acquire a read lock every time. If performance suffers, perhaps 
we can relax that constraint a bit since alerts have a bit more leeway


- Alejandro Fernandez


On March 29, 2016, 7:32 p.m., Jonathan Hurley wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45442/
> ---
> 
> (Updated March 29, 2016, 7:32 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
> Cole.
> 
> 
> Bugs: AMBARI-15620
> https://issues.apache.org/jira/browse/AMBARI-15620
> 
> 
> Repository: ambari
> 
> 
> Description
> ---
> 
> Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after 
> removing a host because they are always considered valid. 
> 
> STR
> - Deploy cluster 
> - Add/Remove nodes a few times 
> - Removed all aded nodes
> 
> {code}
>  There are 4 stale alerts from 4 host(s): 
> amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
> {code}
> 
> 
> Diffs
> -
> 
>   
> ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
>  8dc8e1e 
>   ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
> ebd29e3 
>   ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
> a1ebaba 
>   
> ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
>  6c68d0e 
>   
> ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
>  136a756 
> 
> Diff: https://reviews.apache.org/r/45442/diff/
> 
> 
> Testing
> ---
> 
> mvn clean test
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Re: Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-29 Thread Jonathan Hurley 


> On March 29, 2016, 3:46 p.m., Nate Cole wrote:
> > ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java,
> >  lines 404-405
> > 
> >
> > Would anyone need any other detail?  There's a lot to trigger this: 
> > "Unable to process alert for ... due to ..."   Also, is it really in-error, 
> > or warning?

Good point about error/warning; I think not finding the cluster is probably an 
error. The rest can be warnings. I'll change it. 

As for the message, I'll change it to state a bit more context as to why the 
cluster/host is not valid.


- Jonathan


---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/#review125940
---


On March 29, 2016, 3:32 p.m., Jonathan Hurley wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45442/
> ---
> 
> (Updated March 29, 2016, 3:32 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
> Cole.
> 
> 
> Bugs: AMBARI-15620
> https://issues.apache.org/jira/browse/AMBARI-15620
> 
> 
> Repository: ambari
> 
> 
> Description
> ---
> 
> Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after 
> removing a host because they are always considered valid. 
> 
> STR
> - Deploy cluster 
> - Add/Remove nodes a few times 
> - Removed all aded nodes
> 
> {code}
>  There are 4 stale alerts from 4 host(s): 
> amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
> {code}
> 
> 
> Diffs
> -
> 
>   
> ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
>  8dc8e1e 
>   ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
> ebd29e3 
>   ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
> a1ebaba 
>   
> ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
>  6c68d0e 
>   
> ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
>  136a756 
> 
> Diff: https://reviews.apache.org/r/45442/diff/
> 
> 
> Testing
> ---
> 
> mvn clean test
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Re: Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-29 Thread Nate Cole 

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/#review125940
---


Ship it!





ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
 (lines 404 - 405)


Would anyone need any other detail?  There's a lot to trigger this: "Unable 
to process alert for ... due to ..."   Also, is it really in-error, or warning?



ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
 (lines 392 - 397)


I have no idea how this hostClusterMap relationship came to be :)


- Nate Cole


On March 29, 2016, 3:32 p.m., Jonathan Hurley wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/45442/
> ---
> 
> (Updated March 29, 2016, 3:32 p.m.)
> 
> 
> Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
> Cole.
> 
> 
> Bugs: AMBARI-15620
> https://issues.apache.org/jira/browse/AMBARI-15620
> 
> 
> Repository: ambari
> 
> 
> Description
> ---
> 
> Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after 
> removing a host because they are always considered valid. 
> 
> STR
> - Deploy cluster 
> - Add/Remove nodes a few times 
> - Removed all aded nodes
> 
> {code}
>  There are 4 stale alerts from 4 host(s): 
> amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
> amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
> {code}
> 
> 
> Diffs
> -
> 
>   
> ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
>  8dc8e1e 
>   ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
> ebd29e3 
>   ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
> a1ebaba 
>   
> ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
>  6c68d0e 
>   
> ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
>  136a756 
> 
> Diff: https://reviews.apache.org/r/45442/diff/
> 
> 
> Testing
> ---
> 
> mvn clean test
> 
> 
> Thanks,
> 
> Jonathan Hurley
> 
>

Review Request 45442: Orphaned Host Alerts Cause Stale Alert Notifications After Removing Hosts

2016-03-29 Thread Jonathan Hurley 

---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/45442/
---

Review request for Ambari, Alejandro Fernandez, Dmitro Lisnichenko, and Nate 
Cole.


Bugs: AMBARI-15620
https://issues.apache.org/jira/browse/AMBARI-15620


Repository: ambari


Description
---

Host-level alerts from {{AMBARI}}/{{AMBARI_AGENT}} are orphaned after removing 
a host because they are always considered valid. 

STR
- Deploy cluster 
- Add/Remove nodes a few times 
- Removed all aded nodes

{code}
 There are 4 stale alerts from 4 host(s): 
amb-roll-workflow1458640758-5.novalocal[Host Disk Usage (3h 52m)], 
amb-roll-workflow1458640758-2.novalocal[Host Disk Usage (3h 52m)], 
amb-roll-workflow1458640758-3.novalocal[Host Disk Usage (3h 52m)], 
amb-roll-workflow1458640758-4.novalocal[Host Disk Usage (3h 52m)]
{code}


Diffs
-

  
ambari-server/src/main/java/org/apache/ambari/server/events/listeners/alerts/AlertReceivedListener.java
 8dc8e1e 
  ambari-server/src/main/java/org/apache/ambari/server/orm/dao/HostDAO.java 
ebd29e3 
  ambari-server/src/main/java/org/apache/ambari/server/state/Clusters.java 
a1ebaba 
  
ambari-server/src/main/java/org/apache/ambari/server/state/cluster/ClustersImpl.java
 6c68d0e 
  
ambari-server/src/test/java/org/apache/ambari/server/state/alerts/AlertReceivedListenerTest.java
 136a756 

Diff: https://reviews.apache.org/r/45442/diff/


Testing
---

mvn clean test


Thanks,

Jonathan Hurley