[
https://issues.apache.org/jira/browse/DRILL-6224?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16398522#comment-16398522
]
ASF GitHub Bot commented on DRILL-6224:
---------------------------------------
Github user arina-ielchiieva commented on a diff in the pull request:
https://github.com/apache/drill/pull/1160#discussion_r174445380
--- Diff: exec/java-exec/src/main/resources/rest/index.ftl ---
@@ -192,6 +199,8 @@
var port = getPortNum();
var timeout;
var size = $("#size").html();
+ reloadMemory();
+ setInterval(reloadMemory, refreshTime);
--- End diff --
In graceful shutdown `setTimeout` is used. Which one is better? Anayway
it's better to be consistent in scheduler usage. Plus we do why need two
schedulers? Can one work for both tasks? Plus if one drillbit is unavailable,
there is no need to check memory.
> The metrics' page has gauges reset to near zero values and does not seem to
> update
> ----------------------------------------------------------------------------------
>
> Key: DRILL-6224
> URL: https://issues.apache.org/jira/browse/DRILL-6224
> Project: Apache Drill
> Issue Type: Bug
> Components: Web Server
> Affects Versions: 1.12.0
> Reporter: Kunal Khatua
> Priority: Major
>
> When viewing http://<hostname>:8047/metrics#gauges
> The gauges reset to near zero values and does not seem to update.
> Tracing the server calls made, I see the following:
> {code:json}
> {version: "3.0.0", gauges: {G1-Old-Generation.count: {value: 0},
> G1-Old-Generation.time: {value: 0},…},…}
> counters :
> {drill.connections.rpc.control.encrypted: {count: 0},…}
> gauges :
> {G1-Old-Generation.count: {value: 0}, G1-Old-Generation.time: {value: 0},…}
> histograms : {,…}
> meters : {}
> timers : {}
> version : "3.0.0"
> {code}
> This looks incorrect and would explain why the metrics appear to be
> incomplete.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
