CODEWITHSUNDEEP
azureazure-storageazure-table-storage
Eric Yin
Eric Yin

Reputation: 9003

Azure Storage Table Query, Which is faster?

Very simple lookup, use A to lookup B, A is unique value in whole table, B is unique value in whole table as well

method 1:

PartionKey => A
RowKey => B

lookup by TOP(1) PartionKey eq "A"

method 2:

PartionKey => A
RowKey => "" (empty string)
Value => B

lookup by TOP(1) (PartionKey eq "A") and (RowKey eq "")

method 3:

PartionKey => "" (empty string)
RowKey => A 
Value => B

lookup by TOP(1) (PartionKey eq "") and (RowKey eq "A")

Question: Which is faster?

Upvotes: 1

Views: 1132

Answers (2)

hocho
hocho

Reputation: 1749

If I understand your question… If A is a unique value across the whole table, B’s uniqueness is irrelevant. So you are better off, scalability wise, to split A in two, and use the first part as the Partition Key and the second part as the Row Key.

It is important to understand that in Azure table storage, PartitionKey + RowKey is analogous to the Primary Key in an SQL database, and it’s most efficient to use the whole primary key for a lookup.

Upvotes: 1

user94559
user94559

Reputation: 60153

In method #2, the row key is empty, but you're query specifies a non-empty row key (of "B"). Did you mean and (RowKey eq "")? If so, that method is faster, in that it will always return a result. Method #1 may return an empty result set with a continuation token, at which point you have to make a second query to actually get at the data.

Upvotes: 1

Related Questions