diff --git a/doc/V3HighLevelDesign.md b/doc/V3HighLevelDesign.md
index b0ba3268a7b..5becdb0d46b 100644
--- a/doc/V3HighLevelDesign.md
+++ b/doc/V3HighLevelDesign.md
@@ -782,7 +782,7 @@ Recapitulating what we’ve covered so far:
 
 Once we start allowing joins and subqueries, we have a whole bunch of table aliases and relationships to deal with. We have to contend with name clashes, self-joins, as well as scoping rules. In a way, the vschema has acted as a static symbol table so far. But that’s not going to be enough any more.
 
-The core of the symbol table will contain a map whose key will be a table alias, and the elements will be [similar to the table in vschema](https://github.com/vitessio/vitess/blob/master/go/vt/vtgate/planbuilder/schema.go#L22). However, it will also contain a column list that will be built as the query is parsed.
+The core of the symbol table will contain a map whose key will be a table alias, and the elements will be [similar to the table in vschema](https://github.com/vitessio/vitess/blob/0b3de7c4a2de8daec545f040639b55a835361685/go/vt/vtgate/vindexes/vschema.go#L82). However, it will also contain a column list that will be built as the query is parsed.
 
 ### A simple example
 
diff --git a/doc/VitessQueues.md b/doc/VitessQueues.md
index 8a0ccd2ec67..23a6fd181ce 100644
--- a/doc/VitessQueues.md
+++ b/doc/VitessQueues.md
@@ -79,7 +79,7 @@ capabilities, the usual horizontal sharding process can be used.
 
 Queue Tables are marked in the schema by a comment, in a similar way we detect
 Sequence Tables
-[now](https://github.com/vitessio/vitess/blob/master/go/vt/tabletserver/table_info.go#L37).
+[now](https://github.com/vitessio/vitess/blob/0b3de7c4a2de8daec545f040639b55a835361685/go/vt/vttablet/tabletserver/tabletserver.go#L138).
 
 When a tablet becomes a master, and there are Queue tables, it creates a
 QueueManager for each of them.
diff --git a/java/hadoop/src/main/java/io/vitess/hadoop/README.md b/java/hadoop/src/main/java/io/vitess/hadoop/README.md
index fe88bdc650e..9e4a083bd4c 100644
--- a/java/hadoop/src/main/java/io/vitess/hadoop/README.md
+++ b/java/hadoop/src/main/java/io/vitess/hadoop/README.md
@@ -14,7 +14,7 @@ primary key (id)) Engine=InnoDB;
 
 Let's say we want to write a MapReduce job that imports this table from Vitess to HDFS where each row is turned into a CSV record in HDFS. 
 
-We can use [VitessInputFormat](https://github.com/vitessio/vitess/blob/master/java/hadoop/src/main/java/io/vitess/hadoop/VitessInputFormat.java), an implementation of Hadoop's [InputFormat](https://hadoop.apache.org/docs/stable/api/org/apache/hadoop/mapred/InputFormat.html), for that. With VitessInputFormat, rows from the source table are streamed to the mapper task. Each input record has a [NullWritable](https://hadoop.apache.org/docs/r2.2.0/api/org/apache/hadoop/io/NullWritable.html) key (no key, really), and [RowWritable](https://github.com/vitessio/vitess/blob/master/java/hadoop/src/main/java/io/vitess/hadoop/RowWritable.java) as value, which is a writable implementation for the entire row's contents.
+We can use [VitessInputFormat](https://github.com/vitessio/vitess/blob/master/java/hadoop/src/main/java/io/vitess/hadoop/VitessInputFormat.java), an implementation of Hadoop's [InputFormat](https://hadoop.apache.org/docs/stable/api/org/apache/hadoop/mapred/InputFormat.html), for that. With VitessInputFormat, rows from the source table are streamed to the mapper task. Each input record has a [NullWritable](https://hadoop.apache.org/docs/current/api/org/apache/hadoop/io/NullWritable.html) key (no key, really), and [RowWritable](https://github.com/vitessio/vitess/blob/master/java/hadoop/src/main/java/io/vitess/hadoop/RowWritable.java) as value, which is a writable implementation for the entire row's contents.
 
 Here is an example implementation of our mapper, which transforms each row into a CSV Text.