Kathy Sun has posted comments on this change.

Change subject: draft version -- not completed yet
......................................................................


Patch Set 2:

(19 comments)

http://gerrit.cloudera.org:8080/#/c/3863/1/be/src/exec/info-schema-scan-node.cc
File be/src/exec/info-schema-scan-node.cc:

Line 61:   
ExecEnv::GetInstance()->metrics()->GetChildGroup("impala_metrics")->ToJson(
> I think before we merge this code we'll want to change this so that it gets
I'll leave it for a while. maybe after it could run


Line 122:   int num_ctxs = conjunct_ctxs_.size();
> row_batch->AtCapacity() checks to see if the batch is full, so you are ok h
What I mean here is I assume the batch is smaller than 1024 rows, and I could 
get all of it at one time, instead of get a piece of the whole by 
getNextInputBatch after using out of it.

just simplify question here, I'll expand it after it could run


Line 163: void InfoSchemaScanNode::Close(RuntimeState* state) {
> This can be one line.
Done


Line 167:   PeriodicCounterUpdater::StopRateCounter(total_throughput_counter());
> This can be one line.
Done


Line 174:     stringstream* out) const {
> These DebugString() methods are mainly to help with debugging. I don't thin
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/be/src/runtime/descriptors.h
File be/src/runtime/descriptors.h:

Line 370:   std::string table_name_;
> We'll also want TInfoSchemaTableType here too
I'm not very sure here.
In common/thriftPlanNode.thrift
I add a enum, I considered TInfoSchemaTableType, but I thought it could be 
reasonable to name it as TInfoSchemaTableName, since there is only one table 
each type. 

enum TInfoSchemaTableName {
  impala_metrics,
  datastream_manager,
  impala_server,
  jvm,
  scheduler,
  statestore_subscriber,
}


Did I get you? Or you mean type is not about subgroup in metrics.
Like this?
enum TInfoSchemaTableType {
  metrics,
  catalog,
}
foreshadow for later expansion?


http://gerrit.cloudera.org:8080/#/c/3863/1/common/thrift/CatalogObjects.thrift
File common/thrift/CatalogObjects.thrift:

Line 353: //  // Network address of a master host in the form of 0.0.0.0:port
> I think we'll also need TInfoSchemaTableName here so that we know which tab
If add TInfoSchemaTableName( or Type), it will be the same as table_name, isn't 
it? so there is redundant.

I can replace string table_name with TInfoSchemaTableName...?


PS1, Line 408: Set
> Extra word.
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/common/thrift/PlanNodes.thrift
File common/thrift/PlanNodes.thrift:

Line 21: 
> We should avoid adding extra whitespace like this.
Done


Line 257: 
> Extra line
Done


PS1, Line 260: impala_metrics
> We should use upper case for info schema table names.
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/fe/src/main/java/com/cloudera/impala/catalog/Catalog.java
File fe/src/main/java/com/cloudera/impala/catalog/Catalog.java:

PS1, Line 63: _impala_information_schema
> I think we should just call it "information_schema".
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/fe/src/main/java/com/cloudera/impala/catalog/InformationSchemaDb.java
File fe/src/main/java/com/cloudera/impala/catalog/InformationSchemaDb.java:

Line 1: package com.cloudera.impala.catalog;
> Make sure to include the ASF license headers at the top of all new files (i
Done


PS1, Line 11: 
> Will need to clean up the formatting here.
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/fe/src/main/java/com/cloudera/impala/catalog/InformationSchemaTable.java
File fe/src/main/java/com/cloudera/impala/catalog/InformationSchemaTable.java:

Line 35:       org.apache.hadoop.hive.metastore.api.Table msTable, Db db, 
String name,
> We'll need to make this depend on the type of the information schema table.
Done


http://gerrit.cloudera.org:8080/#/c/3863/1/fe/src/main/java/com/cloudera/impala/planner/InfoSchemaScanNode.java
File fe/src/main/java/com/cloudera/impala/planner/InfoSchemaScanNode.java:

Line 1: // Copyright 2016 Cloudera Inc.
> You'll need to replace all these Cloudera headers with ASF headers. E.g. se
I'll do it in a single patch later, not to mix those with other changes


Line 41:        private final static Logger LOG = LoggerFactory
> Need to fix up these tabs.
fix in next patch as well


PS1, Line 99: eed to change to metrics.size() or estimate
> Good idea. I think it's probably metrics.size() * number of backends?
discuss offline


http://gerrit.cloudera.org:8080/#/c/3863/1/fe/src/main/java/com/cloudera/impala/planner/SingleNodePlanner.java
File fe/src/main/java/com/cloudera/impala/planner/SingleNodePlanner.java:

Line 1328:      } else {
> There are some tabs in here - maybe you need to configure your editor to al
Done


-- 
To view, visit http://gerrit.cloudera.org:8080/3863
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings

Gerrit-MessageType: comment
Gerrit-Change-Id: I7adbeb45220c468e43b424d70c30b952f6cec2cd
Gerrit-PatchSet: 2
Gerrit-Project: Impala-ASF
Gerrit-Branch: master
Gerrit-Owner: Kathy Sun <kathy....@cloudera.com>
Gerrit-Reviewer: Kathy Sun <kathy....@cloudera.com>
Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com>
Gerrit-HasComments: Yes

Reply via email to