[ 
https://issues.apache.org/jira/browse/SENTRY-2106?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Arjun Mishra updated SENTRY-2106:
---------------------------------
    Description: 
After steady state, a full snapshot is triggered by mainly 2 cases
# Sentry is "ahead"
# Sentry is behind
Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable as 
it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG are 
not in sync. As a result of this unnecessary full snapshots can be triggered. 

The solution is to eliminate this check. Sentry can never be ahead of HMS. 

  was:
Once SentryHMSNotification table is not empty, there are two cases when a full 
snapshot is triggered.
# If first event in list of notifications received from HMS greater than latest 
sentry hms notification Id
# If latest sentry notification Id is greater than current hms notification Id

The later is a unexpected case where for some reason sentry gets ahead of HMS. 
We should add a logic to trigger a full snapshot in case 2 only after a 
configurable number of retries. This will avoid  unnecessary full snapshot 
triggers as they are resource intensive


> Eliminate sentry ahead logic
> ----------------------------
>
>                 Key: SENTRY-2106
>                 URL: https://issues.apache.org/jira/browse/SENTRY-2106
>             Project: Sentry
>          Issue Type: New Feature
>          Components: Sentry
>    Affects Versions: 2.1.0
>            Reporter: Arjun Mishra
>            Assignee: Arjun Mishra
>              Labels: features
>             Fix For: 2.1.0
>
>         Attachments: SENTRY-2106.01.patch, SENTRY-2106.02.patch, 
> SENTRY-2106.03.patch
>
>
> After steady state, a full snapshot is triggered by mainly 2 cases
> # Sentry is "ahead"
> # Sentry is behind
> Case 1 has a dependency on NOTIFICATION_SEQUENCE table. This is not reliable 
> as it was observed that sometimes NOTIFICATION_SEQUENCE and NOTIFICATION_LOG 
> are not in sync. As a result of this unnecessary full snapshots can be 
> triggered. 
> The solution is to eliminate this check. Sentry can never be ahead of HMS. 



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to