[
https://issues.apache.org/jira/browse/FALCON-1563?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14986679#comment-14986679
]
Ajay Yadava commented on FALCON-1563:
-------------------------------------
The behaviour explained here is not caused by FALCON-1562 and hence I had
removed the previous link to that JIRA as root cause for this. This issue is
required for 0.8 I believe as it causes serious issues in SLA monitoring under
very common scenarios.
I couldn't get to it over the weekend, I will try to wrap it up as soon as
possible.
> SLA monitoring API does not honour update feature
> -------------------------------------------------
>
> Key: FALCON-1563
> URL: https://issues.apache.org/jira/browse/FALCON-1563
> Project: Falcon
> Issue Type: Bug
> Components: prism
> Affects Versions: 0.8
> Environment: QA
> Reporter: Pragya Mittal
> Assignee: Ajay Yadava
> Priority: Blocker
> Fix For: 0.8
>
>
> When entity definition is updated, SLA monitoring api removes old instances
> and only shows pending instances from now for that entity.
> I submitted a feed with frequency minute(1)
> {noformat}
> <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
> <feed name="sla-feed" description="clicks log" xmlns="uri:falcon:feed:0.1">
> <partitions>
> <partition name="country"/>
> <partition name="colo"/>
> </partitions>
> <frequency>minutes(1)</frequency>
> <sla slaLow="minutes(1)" slaHigh="minutes(3)"/>
> <timezone>UTC</timezone>
> <late-arrival cut-off="hours(6)"/>
> <clusters>
> <cluster name="A9e7e0672-e02bcc65">
> <validity start="2015-10-28T11:58Z" end="2015-10-30T11:17Z"/>
> <retention limit="days(1000000)" action="delete"/>
> </cluster>
> </clusters>
> <locations>
> <location type="data"
> path="/tmp/falcon-regression/FeedSlaMonitoring/input/${YEAR}/${MONTH}/${DAY}/${HOUR}/${MINUTE}"/>
> <location type="stats" path="/projects/falcon/clicksStats"/>
> <location type="meta" path="/projects/falcon/clicksMetaData"/>
> </locations>
> <ACL owner="pragya" group="dataqa" permission="*"/>
> <schema location="/schema/clicks" provider="protobuf"/>
> <properties>
> <property name="field1" value="value1"/>
> </properties>
> </feed>
> {noformat}
> SLA monitoring response :
> {noformat}
> dataqa@lda01:~$ falcon entity -type feed -start 2015-10-26T11:58Z -slaAlert
> name: sla-feed, type: FEED, cluster: A9e7e0672-e02bcc65, instanceTime:
> 2015-10-28T12:01Z, tags: Missed SLA Low
> name: sla-feed, type: FEED, cluster: A9e7e0672-e02bcc65, instanceTime:
> 2015-10-28T12:02Z, tags: Missed SLA Low
> Response: ua1/Success!
> {noformat}
> After updating frequency of feed to 2 minutes response is :
> {noformat}
> dataqa@lda01:~$ falcon entity -type feed -start 2015-10-26T11:58Z -slaAlert
> name: sla-feed, type: FEED, cluster: A9e7e0672-e02bcc65, instanceTime:
> 2015-10-28T12:08Z, tags: Missed SLA Low
> name: sla-feed, type: FEED, cluster: A9e7e0672-e02bcc65, instanceTime:
> 2015-10-28T12:10Z, tags: Missed SLA Low
> Response: ua1/Success!
> Request Id: ua1/1548989490@qtp-2030538903-2 -
> e5d6dd8b-0e1f-44e7-9227-6086d63b6b4f
> {noformat}
> All instances before now(0,0) are deleted from list of pending instances.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
