forked from docs/doc-exports
Yang, Tong
6182f91ba8
Reviewed-by: Hasko, Vladimir <vladimir.hasko@t-systems.com> Co-authored-by: Yang, Tong <yangtong2@huawei.com> Co-committed-by: Yang, Tong <yangtong2@huawei.com>
15 lines
1.4 KiB
HTML
15 lines
1.4 KiB
HTML
<a name="mrs_01_2011"></a><a name="mrs_01_2011"></a>
|
|
|
|
<h1 class="topictitle1">Why Do Executors Fail to be Removed After the NodeManeger Is Shut Down?</h1>
|
|
<div id="body1595920219451"><div class="section" id="mrs_01_2011__s4b5958df25d94966918f3c1cd826dec6"><h4 class="sectiontitle">Question</h4><p id="mrs_01_2011__ad56a0d348da84343958ebb70d1391117">If the NodeManager is shut down with the Executor dynamic allocation enabled, the Executors on the node where the NodeManeger is shut down fail to be removed from the driver page after the idle time expires.</p>
|
|
</div>
|
|
<div class="section" id="mrs_01_2011__s65e168f0b67e4facbd30db02f849d142"><h4 class="sectiontitle">Answer</h4><p id="mrs_01_2011__a07118e07b6644799ad5bbef715cbccac">When the ResourceManager detects that the NodeManager is shut down, the driver has requested to kill Executors due to idle time expiry. However, the Executors cannot actually be killed because the NodeManager is shut down. The driver cannot detect the LOST events of these Executors and does not remove Executors from its Executor list. Therefore, the Executors are not removed from the driver page. This phenomenon is normal after the YARN NodeManager is shut down. The Executors will be removed after the NodeManager restarts.</p>
|
|
</div>
|
|
</div>
|
|
<div>
|
|
<div class="familylinks">
|
|
<div class="parentlink"><strong>Parent topic:</strong> <a href="mrs_01_2003.html">Spark Core</a></div>
|
|
</div>
|
|
</div>
|
|
|