[jira] [Updated] (HDFS-17303) Make WINDOW_SIZE and NUM_WINDOWS configurable.

2024-05-06 Thread Zhaobo Huang (Jira)


 [ 
https://issues.apache.org/jira/browse/HDFS-17303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Zhaobo Huang updated HDFS-17303:

Description: 
1. The delay reported by DN to NN is an average delay of 3 hours.
MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT = 
36
2. Can changeing this window size and num to configurable param?

  was:
1. The delay reported by DN to NN is an average delay of 3 hours, which 
confuses me.
MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT = 
36
2. There is a time limit for SlowNodes collected by nn, which is currently set 
to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the time 
threshold for printing SlowNode logs written downstream is 300ms 
(dfs.datanode.low.io.warning.threshold.ms).
3. Can changeing this window size and num to configurable param?


> Make WINDOW_SIZE and NUM_WINDOWS configurable.
> --
>
> Key: HDFS-17303
> URL: https://issues.apache.org/jira/browse/HDFS-17303
> Project: Hadoop HDFS
>  Issue Type: Improvement
>Reporter: Zhaobo Huang
>Assignee: Zhaobo Huang
>Priority: Major
>  Labels: pull-request-available
>
> 1. The delay reported by DN to NN is an average delay of 3 hours.
> MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT 
> = 36
> 2. Can changeing this window size and num to configurable param?



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org



[jira] [Updated] (HDFS-17303) Make WINDOW_SIZE and NUM_WINDOWS configurable.

2024-05-06 Thread ASF GitHub Bot (Jira)


 [ 
https://issues.apache.org/jira/browse/HDFS-17303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

ASF GitHub Bot updated HDFS-17303:
--
Labels: pull-request-available  (was: )

> Make WINDOW_SIZE and NUM_WINDOWS configurable.
> --
>
> Key: HDFS-17303
> URL: https://issues.apache.org/jira/browse/HDFS-17303
> Project: Hadoop HDFS
>  Issue Type: Improvement
>Reporter: Zhaobo Huang
>Assignee: Zhaobo Huang
>Priority: Major
>  Labels: pull-request-available
>
> 1. The delay reported by DN to NN is an average delay of 3 hours, which 
> confuses me.
> MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT 
> = 36
> 2. There is a time limit for SlowNodes collected by nn, which is currently 
> set to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the 
> time threshold for printing SlowNode logs written downstream is 300ms 
> (dfs.datanode.low.io.warning.threshold.ms).
> 3. Can changeing this window size and num to configurable param?



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org



[jira] [Updated] (HDFS-17303) Make WINDOW_SIZE and NUM_WINDOWS configurable.

2023-12-25 Thread huangzhaobo99 (Jira)


 [ 
https://issues.apache.org/jira/browse/HDFS-17303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

huangzhaobo99 updated HDFS-17303:
-
Description: 
1. The delay reported by DN to NN is an average delay of 3 hours, which 
confuses me.
MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT = 
36
2. There is a time limit for SlowNodes collected by nn, which is currently set 
to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the time 
threshold for printing SlowNode logs written downstream is 300ms 
(dfs.datanode.low.io.warning.threshold.ms).
3. Can changeing this window size and num to configurable param?

  was:
1. The delay reported by DN to NN is an average delay of 3 hours, which 
confuses me.
2. There is a time limit for SlowNodes collected by nn, which is currently set 
to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the time 
threshold for printing SlowNode logs written downstream is 300ms 
(dfs.datanode.low.io.warning.threshold.ms).
3. Can changeing this window size and num to configurable param?


> Make WINDOW_SIZE and NUM_WINDOWS configurable.
> --
>
> Key: HDFS-17303
> URL: https://issues.apache.org/jira/browse/HDFS-17303
> Project: Hadoop HDFS
>  Issue Type: Improvement
>Reporter: huangzhaobo99
>Priority: Major
>
> 1. The delay reported by DN to NN is an average delay of 3 hours, which 
> confuses me.
> MutableRollingAverages: WINDOW_SIZE_MS_DEFAULT = 300_000 NUM_WINDOWS_DEFAULT 
> = 36
> 2. There is a time limit for SlowNodes collected by nn, which is currently 
> set to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the 
> time threshold for printing SlowNode logs written downstream is 300ms 
> (dfs.datanode.low.io.warning.threshold.ms).
> 3. Can changeing this window size and num to configurable param?



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org



[jira] [Updated] (HDFS-17303) Make WINDOW_SIZE and NUM_WINDOWS configurable.

2023-12-25 Thread huangzhaobo99 (Jira)


 [ 
https://issues.apache.org/jira/browse/HDFS-17303?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

huangzhaobo99 updated HDFS-17303:
-
Description: 
1. The delay reported by DN to NN is an average delay of 3 hours, which 
confuses me.
2. There is a time limit for SlowNodes collected by nn, which is currently set 
to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the time 
threshold for printing SlowNode logs written downstream is 300ms 
(dfs.datanode.low.io.warning.threshold.ms).
3. Can changeing this window size and num to configurable param?

> Make WINDOW_SIZE and NUM_WINDOWS configurable.
> --
>
> Key: HDFS-17303
> URL: https://issues.apache.org/jira/browse/HDFS-17303
> Project: Hadoop HDFS
>  Issue Type: Improvement
>Reporter: huangzhaobo99
>Priority: Major
>
> 1. The delay reported by DN to NN is an average delay of 3 hours, which 
> confuses me.
> 2. There is a time limit for SlowNodes collected by nn, which is currently 
> set to 5ms by default (dfs.datanode.slowpeer.low.threshold.ms), while the 
> time threshold for printing SlowNode logs written downstream is 300ms 
> (dfs.datanode.low.io.warning.threshold.ms).
> 3. Can changeing this window size and num to configurable param?



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

-
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org