[GitHub] [incubator-pinot] sunithabeeram commented on issue #4380: Define SLO config for tables
sunithabeeram commented on issue #4380: Define SLO config for tables URL: https://github.com/apache/incubator-pinot/pull/4380#issuecomment-521829184 Closing this as this has been moved to higher layers. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
[GitHub] [incubator-pinot] sunithabeeram commented on issue #4380: Define SLO config for tables
sunithabeeram commented on issue #4380: Define SLO config for tables URL: https://github.com/apache/incubator-pinot/pull/4380#issuecomment-511248076 @kishoreg sorry about the delay in responses on my end as I am traveling. Can you elaborate on what you mean by guaranteeing the latency numbers? Unless these are measured, there isn't a way to guarantee them, right? Once the numbers are available, Pinot operators can make appropriate changes to improve the system. Client side measurements include noise (network and other client side induced latency) and are not reliable measures. The config helps emit the base metrics that are can be reliably used in measuring availability numbers from server side. Many services at LinkedIn use server side measurements - we have been emitting these currently from the broker using Jetty/container configs - but they make per table metrics harder to compute (as broker multi-tenancy is common). This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
[GitHub] [incubator-pinot] sunithabeeram commented on issue #4380: Define SLO config for tables
sunithabeeram commented on issue #4380: Define SLO config for tables URL: https://github.com/apache/incubator-pinot/pull/4380#issuecomment-509900678 @kishoreg - do you have further thoughts/comments on this? This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
[GitHub] [incubator-pinot] sunithabeeram commented on issue #4380: Define SLO config for tables
sunithabeeram commented on issue #4380: Define SLO config for tables URL: https://github.com/apache/incubator-pinot/pull/4380#issuecomment-508666033 @kishoreg these are required in the table config as they are per table parameters. The broker emits metrics that track against these numbers. There are no guarantees around this - they are used to obtain overall availability numbers of the service at the desired threshold (ie, 99.999/99.99 etc). The time interval over which the numbers are derived is not dictated by this implementation - its left to the users to derive. This change and the ones in #4356 emit a per response healthy/unhealthy score. This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services - To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org