Add Create An Index Without Locking The Table as a postgres til

README.md

@@ -10,7 +10,7 @@ smart people at [Hashrocket](http://hashrocket.com/).
 For a steady stream of TILs from a variety of rocketeers, checkout
 [til.hashrocket.com](https://til.hashrocket.com/).
 
-_817 TILs and counting..._
+_818 TILs and counting..._
 
 ---
 
@@ -393,6 +393,7 @@ _817 TILs and counting..._
 - [Convert A String To A Timestamp](postgres/convert-a-string-to-a-timestamp.md)
 - [Count Records By Type](postgres/count-records-by-type.md)
 - [Create A Composite Primary Key](postgres/create-a-composite-primary-key.md)
+- [Create An Index Without Locking The Table](postgres/create-an-index-without-locking-the-table.md)
 - [Create hstore From Two Arrays](postgres/create-hstore-from-two-arrays.md)
 - [Create Table Adds A Data Type](postgres/create-table-adds-a-data-type.md)
 - [Creating Conditional Constraints](postgres/creating-conditional-constraints.md)

postgres/create-an-index-without-locking-the-table.md

@@ -0,0 +1,20 @@
+# Create An Index Without Locking The Table
+
+When creating an index for a column, the process of building the index will
+lock the column's table. For small datasets this isn't a concern because the
+index will take no time at all to create. For larger datasets, the lock could
+last long enough to create meaningful downtime. This can all be avoided by
+telling Postgres to build the index concurrently.
+
+```sql
+create index concurrently idx_book_isbns on books(isbn);
+```
+
+Creating the index this way will take a bit longer and put more strain on
+machine resources, but it allows concurrent inserts, updates, or deletes on the
+table. In other words, you can add an index to a large table in a production
+environment without bringing down your app.
+
+Read more about the [details and potential
+caveats](https://www.postgresql.org/docs/current/sql-createindex.html#SQL-CREATEINDEX-CONCURRENTLY)
+in the docs.