[ 
https://issues.apache.org/jira/browse/IGNITE-12844?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Philipp Masharov updated IGNITE-12844:
--------------------------------------
    Description: 
Several thousand caches send metrics in each discovery message.
 Check implementation of `TcpDiscoveryMetricsUpdateMessage`
 This behavior increases processing time and raises message size. As a result, 
the discovery worker is busy and the queue pool is growing.
 We need to notify about the possible issue and print a suggestion, for 
example, `-DIGNITE_DISCOVERY_DISABLE_CACHE_METRICS_UPDATE=true` if it can help. 
Assumes to print WARN message once, for example, if 
TcpDiscoveryMetricsUpdateMessage consists of more than 500 CacheMetricsSnaphot 
instances.

  was:
Several thousand caches send metrics in each discovery message.
Check implementation of `TcpDiscoveryMetricsUpdateMessage`
This behavior increases processing time and raises message size. As a result, 
the discovery worker is busy and the queue pool is growing.
We need to notify about the possible issue and print suggestion, for example, 
`-DIGNITE_DISCOVERY_DISABLE_CACHE=true` if it can help. Assumes to print WARN 
message once, for example, if TcpDiscoveryMetricsUpdateMessage consists of more 
than 500 CacheMetricsSnaphot instances.


> Too many caches can block discovery
> -----------------------------------
>
>                 Key: IGNITE-12844
>                 URL: https://issues.apache.org/jira/browse/IGNITE-12844
>             Project: Ignite
>          Issue Type: Bug
>            Reporter: Philipp Masharov
>            Assignee: Philipp Masharov
>            Priority: Major
>             Fix For: 2.9
>
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> Several thousand caches send metrics in each discovery message.
>  Check implementation of `TcpDiscoveryMetricsUpdateMessage`
>  This behavior increases processing time and raises message size. As a 
> result, the discovery worker is busy and the queue pool is growing.
>  We need to notify about the possible issue and print a suggestion, for 
> example, `-DIGNITE_DISCOVERY_DISABLE_CACHE_METRICS_UPDATE=true` if it can 
> help. Assumes to print WARN message once, for example, if 
> TcpDiscoveryMetricsUpdateMessage consists of more than 500 
> CacheMetricsSnaphot instances.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

Reply via email to