[
https://issues.apache.org/jira/browse/MAPREDUCE-7100?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Xiang Li updated MAPREDUCE-7100:
--------------------------------
Description:
We are using hadoop 2.7.3 and the computing layer is running out of the storage
cluster (that is, node managers are running on a different set of nodes from
data node). The problem we meet is that the container allocation is quite slow.
After some debugging, we found that in
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor#addContainerReq()
(the following code is from trunk, not 2.7.3)
{code}
protected void addContainerReq(ContainerRequest req) {
// Create resource requests
for (String host : req.hosts) {
// Data-local
if (!isNodeBlacklisted(host)) {
addResourceRequest(req.priority, host, req.capability,
null);
}
}
// Nothing Rack-local for now
for (String rack : req.racks) {
addResourceRequest(req.priority, rack, req.capability,
null);
}
// Off-switch
addResourceRequest(req.priority, ResourceRequest.ANY, req.capability,
req.nodeLabelExpression);
}
{code}
The request of data-local and rack-local could be skipped when computing layer
is not the same as the storage cluster.
If I get it correctly, req.hosts and req.racks are provided by InputFormat. If
the mapper is to read HDFS, req.hosts is the corresponding data node and
req.racks is its rack. The debug log of AM is like:
{code}
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor: addResourceRequest:
applicationId=1 priority=20 resourceName=<data-node> numContainers=1…256 #asks=1
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor: addResourceRequest:
applicationId=1 priority=20 resourceName=<its rack> numContainers=1…256 #asks=2
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor: addResourceRequest:
applicationId=1 priority=20 resourceName=* numContainers=1…256 #asks=3
{code}
Although eventually, the resource request with resourceName=<data-node> will
not be satisfied (because the data node is not node manager), it could be
better that if we know that computing layer is not the same as the storage
cluster, the request of data-node and rack-local could be skipped (by options)
in a earlier stage.
was:
We are using hadoop 2.7.3 and the computer layer is running out of the storage
cluster (that is, node managers are running on a different set of nodes from
data node). The problem we meet is that the container allocation is quite slow.
After some debugging, we found that in
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor#addContainerReq()
{code}
protected void addContainerReq(ContainerRequest req) {
// Create resource requests
for (String host : req.hosts) {
// Data-local
if (!isNodeBlacklisted(host)) {
addResourceRequest(req.priority, host, req.capability,
null);
}
}
// Nothing Rack-local for now
for (String rack : req.racks) {
addResourceRequest(req.priority, rack, req.capability,
null);
}
// Off-switch
addResourceRequest(req.priority, ResourceRequest.ANY, req.capability,
req.nodeLabelExpression);
}
{code}
> Provide options to skip adding container request for data-local and
> rack-local respectively
> -------------------------------------------------------------------------------------------
>
> Key: MAPREDUCE-7100
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-7100
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: applicationmaster
> Reporter: Xiang Li
> Priority: Minor
>
> We are using hadoop 2.7.3 and the computing layer is running out of the
> storage cluster (that is, node managers are running on a different set of
> nodes from data node). The problem we meet is that the container allocation
> is quite slow.
> After some debugging, we found that in
> org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor#addContainerReq()
> (the following code is from trunk, not 2.7.3)
> {code}
> protected void addContainerReq(ContainerRequest req) {
> // Create resource requests
> for (String host : req.hosts) {
> // Data-local
> if (!isNodeBlacklisted(host)) {
> addResourceRequest(req.priority, host, req.capability,
> null);
> }
> }
> // Nothing Rack-local for now
> for (String rack : req.racks) {
> addResourceRequest(req.priority, rack, req.capability,
> null);
> }
> // Off-switch
> addResourceRequest(req.priority, ResourceRequest.ANY, req.capability,
> req.nodeLabelExpression);
> }
> {code}
> The request of data-local and rack-local could be skipped when computing
> layer is not the same as the storage cluster.
> If I get it correctly, req.hosts and req.racks are provided by InputFormat.
> If the mapper is to read HDFS, req.hosts is the corresponding data node and
> req.racks is its rack. The debug log of AM is like:
> {code}
> org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor:
> addResourceRequest: applicationId=1 priority=20 resourceName=<data-node>
> numContainers=1…256 #asks=1
> org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor:
> addResourceRequest: applicationId=1 priority=20 resourceName=<its rack>
> numContainers=1…256 #asks=2
> org.apache.hadoop.mapreduce.v2.app.rm.RMContainerRequestor:
> addResourceRequest: applicationId=1 priority=20 resourceName=*
> numContainers=1…256 #asks=3
> {code}
> Although eventually, the resource request with resourceName=<data-node> will
> not be satisfied (because the data node is not node manager), it could be
> better that if we know that computing layer is not the same as the storage
> cluster, the request of data-node and rack-local could be skipped (by
> options) in a earlier stage.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]