When defining a Graph Database in CosmosDB a Partition Key must be specified. The Partition Key is used for sharding the database. Each partition has a hard storage limit of 10GB. As such queries that do writes or reads across partitions are a lot more expensive. I want to use CosmosDb to find mutual acquaintances that a user has in common in order to suggest possible connections for them to follow. This is in a social network pattern like Twitter or Instagram where the relationships are of the type of "follower" not bi-directional "friends" like Facebook.
The only data that will be stored in the graph is:
User with 1 property - ID (Vertex)
Relationship - Follows (Edge)
The queries that will be run against the database are:
Create User
Delete User
Find possible mutual acquaintances
What is an appropriate Partition Key/Structure for a graph problem like this?
Any property to be considered as partition key should have large number of unique values and in social network type scenarios you want to keep user generated content for a particular user in his partition. So ID of the user (a GUID) can be a good partition key.
Related
Is there some operation of the Scan API or the Query API that allows to perform a lookup on a table with a composite key (pk/sk) but that varies only in the pk to optimize the Scan operation of the table ?
Let me introduce a use case:
Suppose I have a partition key defined by the id of a project and within each project I have a huge amount of records (sk)
Now, I need to solve the query "return all projects". So I don't have a partition key and I have to perform a scan.
I know that I could create a GSI that solves this problem, but let's assume that this is not the case.
Is there any way to perform a scan that "hops" between each pk, ignoring the elements of the sk's?
In other words, I will collect the information of the first record of each partition key.
DynamoDB is a NoSQL database, as you already know. It is optimized for LOOKUP, and practices that you used to have in SQL databases or other (low-scale) databases are not always available in DynamoDB.
The concept of a partition key is to put records that are part of the same partition together and sorted by the sort key. The other side of it is that records that don't have the same partition key, are stored in other locations. It is not a long list (or tree) of records that you can scan over.
When you design your schema in a NoSQL database, you need to consider the access pattern to that data. If you need a list of all the projects, you need to maintain an index that will allow it.
I'm trying to create my first DynamoDB based project and I'm having some trouble figuring out the best practices working with a NoSQL database.
My usecase currently is storing users and teams. I have a table that has a partition key of either USER#{userId} or TEAM{#teamId}. If the PK is TEAM{#teamId} I store records with SK either TEAM#{teamId} for team details, or USER#{userId} for the user's details in the team (acceptedInvite, joinDate etc). I also have a GSI based on the userId/email column that allows me to query all the teams a user has been invted to, or the user's team, depending on the value of acceptedInvite field. Attached screenshots of the table structure at the moment:
The table
The GSI
In my application I have an access pattern of getting a team's team members, given a user id.
Currently, I'm doing two queries in my lambda function:
Get user's team, by querying the GSI on PK = {userId} and fitler acceptedInvite = true
Get the team data by querying the table on PK = {teamId} and SK begins_with USER#
This works fine, but I'm concerned I need to preform two separate DynamoDB calls in my API function.
I'm wondering if there's a better way to represent this access pattern and if multiple dynamoDB calls are actually that bad, since I cannot see another way to do this.
Any kind of feedback is appreciated!
The best way to avoid making two queries like this is to supply the API caller with all the information needed to make a single DynamoDB request. For your case this means supplying the caller with the teamId. You can do this as either as part of a list operation response, or if it is the authenticated user, then as part of their claims in a JWT.
I have an application on AWS using DynamoDB with user sending messages to each other. I am not familiar with AWS and I a lacking best practice knowledge
My application has now started to get slow to retrieve messages for a user because I have more and more data in my database.
I am thinking that it is because of my primary key and I wonder what could be a good primary key in this case.
Currently I am using a random guid as a primary key.
I am looking to retrieve all messages corresponding to a user, I am doing a scan operation.
I would like to use a composite value based on username as a primary key but I wonder if it will be better. For instance if I need to retrieve the number of messages for a user and to increment it will probably be even longer to do the request to create the primary key.
What would be a good primary key here ?
Thanks!
It will be better since it appears you often query based on the userid. Scans are expensive and should be avoided where possible. AWS has a great article on best practices for choosing a partition key (primary key). The key takeaway is the following:
You should evaluate various approaches based on your data ingestion and access pattern, then choose the most appropriate key with the least probability of hitting throttling issues.
Using a guid for the partition/primary key is a waste if you never query the data using it. Since using the query operation (rather than using scan) requires querying using the partition/primary (and sort key), you want to ensure you choose a value that you use to retrieve the data often and also has the sufficient cardinality to ensure your data is distributed across a reasonable amount of partitions.
What other access patterns do you have in your application? From what you've mentioned so far, userid seems to be a reasonable choice.
We are using cosmos db for our data storage and there is a case where I have to do cross partition query because I don't know the specific partition key. But I will know a part of it.
To elaborate, my partition key is combination of multiple strings, lets say A-B.
and lets say I only know A but not B. So is there any way to do wild card searching on the partition key.
would that optimize the query or its not possible. Depending on that I will consider if to put A in the the partition key at all or not
Based on my researching and Partitioning in Azure Cosmos DB, nowhere mentions cosmos db partition key supports wildcard searching feature. Only index policy supports wildcard setting:https://learn.microsoft.com/en-us/azure/cosmos-db/index-policy#including-and-excluding-property-paths
So,for your situation,you don't know B so that i'd suggest you considering setting pk as A. Besides,you could vote up this thread:https://github.com/PlagueHO/CosmosDB/issues/153
I've been reading some DynamoDB index docs and they've left me more confused than anything. Let's clear the air with a concrete example.
I have a simple calendar application, where I have an events table. Here are the columns I have:
id: guid,
name: string,
startTimestamp: integer,
calendarId: guid (foreign key in a traditional RDBMS model)
ownerId: guid (foreign key in a traditional RDBMS model)
I'd like to perform queries such as:
Get an event by ID
Get all events where calendarId = x and ownerId = y
Get all events where startTimestamp is between x and y and calendarId = z
DynamoDB docs seem to heavily suggest avoiding using the event's ID as a partition/sort key here, so what's the recommended schema?
This is a problem that everyone wrestles with when they start with (and indeed when they are experienced with) DynamoDB.
Pricing and throughput
Let's start with how DynamoDB is priced (its related - honestly). Ignoring the free tier for a moment, you pay $0.25 per GB per month for data at rest. You also pay $0.47 per Write Capacity Unit (WCU) per month and $0.09 per Read Capacity Unit (RCU) per month. Throughput is the number of WCUs and RCUs on your table. You have to specify throughput up front on your table - the volume of writes and reads you can perform on your table is limited by your throughput provision. Pay more money and you can do more reads and writes per second. The exact details of how DynamoDB partitions tables can be found in this answer.
Keys
Now we need to consider table partitioning. Tables must have a primary key. A primary key must have a hash key (aka a partition key) and may optionally have a sort key (aka a range key). DynamoDB creates partitions based on your hash key values. Within a partition key value the data is sorted by range key, if you have specified one.
Data Access
If you have the exact primary key (hash key and range key if there is one), you can instantly access an item using GetItem. If you have multiple items to get, you can use BatchGetItem.
DynamoDB can only 'search' data in two ways. A Query can only take data from one partition in one call, because it uses the partition key (and optionally a sort key) it is quick. A Scan always evaluates every item in table, so its typically slow and doesn't scale well on large tables.
Throughput distribution
This is where is gets interesting. DynamoDB takes all the throughput you have purchased and evenly spreads it over all of you table partitions. Imagine you have 10 WCUs and 10 RCUs on your table, and 5 partitions, that means you have 2 WCUs and 2 RCUs per partition. That's fine if you access each partition evenly, you get to use all of your purchased throughput. But imagine you only ever access one partition. Now you've purchased 10 WCUs and RCUs but you are only using 2. Your table is going to be much slower than you thought. One option is to just buy more throughput, that will work, but its probably not very satisfactory to most engineers.
Uniform Access v Natural Access
Based on the above we know we want to design a table where each partition gets accessed evenly. However, in my experience people get too hung up about this, which is not surprising if you read the article I just linked (which you also linked).
Remember that partition keys is what we use in a Query to get our data fast, and avoid regular Scans. Some people get too focussed making their partition access perfectly uniform, and end up with a table they can't query quickly.
The answer
I like to refer to Best Practices for Tables guide. And particularly the table where it says User ID is a good partition key so long many user access your application regularly. (It actually says where you have many users - which is not correct, the size of the table is irrelevant).
Its a balance between uniform access and being able to use intuitive, natural queries for your application, but what I am saying is, if you are new to DyanmoDB, the right answer probably is to design your table based on intuitive access. After you've done that successfully, have a think about uniform access and hot partitions, but just remember access doesn't have to be perfectly uniform. There are various design patterns to achieve both intuitive and uniform access, but these can be complicated for those starting out and in many cases can probably discourage people using DynamoDB if they get too focussed on the uniform access idea.
Tips
Most applications will have users. For most queries, in most applications, the most common query you will do is get data for a user. So the first option for most application's primary partition key will often be a user id. That's fine, as long as you don't have a few very high hitting users and many users that never log in.
Another tip. If your table is called vegetables, your primary partition key will probably be vegetable id. If your table is called shoes, your primary partition key will probably be shoe id.
Most applications will have many items for each user (or vegetable or shoe). The primary key has to be unique. A good option often is to add a date range (sort) key - perhaps the datetime the item was created. This then orders the items within the user partition by creation date, and also gives each item a unique composite primary key (i.e. hash key + range key). It's also fine to use a generated UUID as a range key, you wont use the ordering it gives you, but you can then have many items per user and still use the Query function.
Indexes are not a solution
Aha! But I can just make my partition key totally random, then apply an index with a partition key of the attribute I really want to query on. That way I get uniform access AND fast intutive queries.
Sadly not. Indexes have their own throughput and partitioning, separate to the table the index is built on. Just imagine indexes as a whole new table - that's basically what they are. Indexes are not a work around to uneven partition access.
Finally - your schema
Primary Key
Hash Key: Event ID
Range Key: None
Global Secondary index
Hash Key: Calendar ID
Range Key: startTimestamp
Assuming Event ID is uniformly accessed, it would be a great hash key. You would really need to describe how your data is distributed to discuss this much more. Other things that come in to play are how fast you want queries to work and how much you are willing to pay (e.g. secondary indexes are expensive).
And your queries:
Get an event by ID
GetItem using Event ID
Get all events where calendarId = x and ownerId = y
Query by GSI parition key, add a condition on ownerId
Get all events where startTimestamp is between x and y and calendarId = z
Query by GSI parition key, add a condition on range key
I just want to add something to the accepted anwser:
Get all events where calendarId = x and ownerId = y
Query by GSI parition key, add a condition on ownerId
This method is not reliable. I guess that when you say "add a condition on ownerId", you mean "add a Filter expression on ownerId" (Definition by Alex DeBrie)
But the 1MB read limit by DynamoDB makes it unreliable.
It is better explained in the link above, but here is the sumup:
If you calendar has a lot of events, that represent data with size over 1MB, the results on which you apply the condition ownerId==X will be truncated to the first 1MB, excluding the rest of the data.