Build fault tolerant applications with Cassandra API for Azure Cosmos DB

Azure Cosmos DB is a resource governed system that allows you to execute a certain number of operations per second based on the provisioned throughput you have configured. If clients exceed that limit and consume more request units than what was provisioned, it leads to rate limiting of subsequent requests and exceptions being thrown – they are also referred to as 429 errors.

With the help of a practical example, I’ll demonstrate how to incorporate fault-tolerance in your Go applications by handling and retrying operations affected by these rate limiting errors. To help you follow along, the sample application code for this blog is available on GitHub and it uses the gocql driver for Apache Cassandra. In this post, we’ll go through:

• Initial setup and configuration before running the sample application
• Execution of various load test scenarios and analyze the results
• A quick overview of the Retry Policy implementation.

One way of tackling rate limiting is by adjusting provisioned throughput to meet your application requirements. There are multiple ways to do this, including using Azure portal, Azure CLI, and CQL (Cassandra Query Language) commands.

But, what if you wanted to handle these errors in the application itself?

The good thing is that the Cassandra API for Azure Cosmos DB translates the rate limiting exceptions into overloaded errors on the Cassandra native protocol. Since the gocql driver allows you to plugin your own RetryPolicy, you can write a custom implementation to intercept these errors and retry them after a certain (cool down) time period. This policy can then be applied to each Query or at a global level using a ClusterConfig.

The Azure Cosmos DB extension library makes it quite easy to use Retry Policies in your Java applications. An equivalent Go version is available on GitHub and has been used in the sample application for this blog post.

Retry Policy in action

As promised, you will walk through the entire process using a simple yet practical example. The sample application used to demonstrate the concepts is a service that exposes a REST endpoint to POST orders data which is persisted to a Cassandra table in Azure Cosmos DB.

You will run a few load tests on this API service to see how rate limiting manifests itself and how it’s handled.

Pre-requisites

Start by installing hey, a load testing program. You can download OS specific binaries (64-bit) for Linux, Mac and Windows (please refer to the GitHub repo for latest information in case you face issues downloading the utility)

You can use any other tool that allows you to generate load on an HTTP endpoint

Clone this GitHub repo and change into the right directory:

git clone github.com/abhirockzz/cosmos-go-rate-limiting 
cd cosmos-go-rate-limiting

#cassandra api #apache cassandra #appdev #cassandra #go #paas

What is GEEK

Buddha Community

Build Fault Tolerant Applications With Cassandra API for Azure Cosmos DB

Azure Cosmos DB is a resource governed system that allows you to execute a certain number of operations per second based on the provisioned throughput you have configured. If clients exceed that limit and consume more request units than what was provisioned, it leads to rate limiting of subsequent requests and exceptions being thrown — they are also referred to as 429 errors.

With the help of a practical example, I’ll demonstrate how to incorporate fault-tolerance in your Go applications by handling and retrying operations affected by these rate limiting errors. To help you follow along, the sample application code for this blog is available on GitHub — it uses the gocql driver for Apache Cassandra.

#database #tutorial #nosql #azure #cassandra #azure cosmos db

Learn NoSQL in Azure: Diving Deeper into Azure Cosmos DB

This article is a part of the series – Learn NoSQL in Azure where we explore Azure Cosmos DB as a part of the non-relational database system used widely for a variety of applications. Azure Cosmos DB is a part of Microsoft’s serverless databases on Azure which is highly scalable and distributed across all locations that run on Azure. It is offered as a platform as a service (PAAS) from Azure and you can develop databases that have a very high throughput and very low latency. Using Azure Cosmos DB, customers can replicate their data across multiple locations across the globe and also across multiple locations within the same region. This makes Cosmos DB a highly available database service with almost 99.999% availability for reads and writes for multi-region modes and almost 99.99% availability for single-region modes.

In this article, we will focus more on how Azure Cosmos DB works behind the scenes and how can you get started with it using the Azure Portal. We will also explore how Cosmos DB is priced and understand the pricing model in detail.

How Azure Cosmos DB works

As already mentioned, Azure Cosmos DB is a multi-modal NoSQL database service that is geographically distributed across multiple Azure locations. This helps customers to deploy the databases across multiple locations around the globe. This is beneficial as it helps to reduce the read latency when the users use the application.

As you can see in the figure above, Azure Cosmos DB is distributed across the globe. Let’s suppose you have a web application that is hosted in India. In that case, the NoSQL database in India will be considered as the master database for writes and all the other databases can be considered as a read replicas. Whenever new data is generated, it is written to the database in India first and then it is synchronized with the other databases.

Consistency Levels

While maintaining data over multiple regions, the most common challenge is the latency as when the data is made available to the other databases. For example, when data is written to the database in India, users from India will be able to see that data sooner than users from the US. This is due to the latency in synchronization between the two regions. In order to overcome this, there are a few modes that customers can choose from and define how often or how soon they want their data to be made available in the other regions. Azure Cosmos DB offers five levels of consistency which are as follows:

• Strong
• Bounded staleness
• Session
• Consistent prefix
• Eventual

In most common NoSQL databases, there are only two levels – Strong and Eventual. Strong being the most consistent level while Eventual is the least. However, as we move from Strong to Eventual, consistency decreases but availability and throughput increase. This is a trade-off that customers need to decide based on the criticality of their applications. If you want to read in more detail about the consistency levels, the official guide from Microsoft is the easiest to understand. You can refer to it here.

Azure Cosmos DB Pricing Model

Now that we have some idea about working with the NoSQL database – Azure Cosmos DB on Azure, let us try to understand how the database is priced. In order to work with any cloud-based services, it is essential that you have a sound knowledge of how the services are charged, otherwise, you might end up paying something much higher than your expectations.

If you browse to the pricing page of Azure Cosmos DB, you can see that there are two modes in which the database services are billed.

• Database Operations – Whenever you execute or run queries against your NoSQL database, there are some resources being used. Azure terms these usages in terms of Request Units or RU. The amount of RU consumed per second is aggregated and billed
• Consumed Storage – As you start storing data in your database, it will take up some space in order to store that data. This storage is billed per the standard SSD-based storage across any Azure locations globally

Let’s learn about this in more detail.

#azure #azure cosmos db #nosql #azure #nosql in azure #azure cosmos db

Build fault tolerant applications with Cassandra API for Azure Cosmos DB

Azure Cosmos DB is a resource governed system that allows you to execute a certain number of operations per second based on the provisioned throughput you have configured. If clients exceed that limit and consume more request units than what was provisioned, it leads to rate limiting of subsequent requests and exceptions being thrown – they are also referred to as 429 errors.

With the help of a practical example, I’ll demonstrate how to incorporate fault-tolerance in your Go applications by handling and retrying operations affected by these rate limiting errors. To help you follow along, the sample application code for this blog is available on GitHub and it uses the gocql driver for Apache Cassandra. In this post, we’ll go through:

• Initial setup and configuration before running the sample application
• Execution of various load test scenarios and analyze the results
• A quick overview of the Retry Policy implementation.

One way of tackling rate limiting is by adjusting provisioned throughput to meet your application requirements. There are multiple ways to do this, including using Azure portal, Azure CLI, and CQL (Cassandra Query Language) commands.

But, what if you wanted to handle these errors in the application itself?

The good thing is that the Cassandra API for Azure Cosmos DB translates the rate limiting exceptions into overloaded errors on the Cassandra native protocol. Since the gocql driver allows you to plugin your own RetryPolicy, you can write a custom implementation to intercept these errors and retry them after a certain (cool down) time period. This policy can then be applied to each Query or at a global level using a ClusterConfig.

The Azure Cosmos DB extension library makes it quite easy to use Retry Policies in your Java applications. An equivalent Go version is available on GitHub and has been used in the sample application for this blog post.

Retry Policy in action

As promised, you will walk through the entire process using a simple yet practical example. The sample application used to demonstrate the concepts is a service that exposes a REST endpoint to POST orders data which is persisted to a Cassandra table in Azure Cosmos DB.

You will run a few load tests on this API service to see how rate limiting manifests itself and how it’s handled.

Pre-requisites

Start by installing hey, a load testing program. You can download OS specific binaries (64-bit) for Linux, Mac and Windows (please refer to the GitHub repo for latest information in case you face issues downloading the utility)

You can use any other tool that allows you to generate load on an HTTP endpoint

Clone this GitHub repo and change into the right directory:

git clone github.com/abhirockzz/cosmos-go-rate-limiting 
cd cosmos-go-rate-limiting

#cassandra api #apache cassandra #appdev #cassandra #go #paas

Top 10 API Security Threats Every API Team Should Know

As more and more data is exposed via APIs either as API-first companies or for the explosion of single page apps/JAMStack, API security can no longer be an afterthought. The hard part about APIs is that it provides direct access to large amounts of data while bypassing browser precautions. Instead of worrying about SQL injection and XSS issues, you should be concerned about the bad actor who was able to paginate through all your customer records and their data.

Typical prevention mechanisms like Captchas and browser fingerprinting won’t work since APIs by design need to handle a very large number of API accesses even by a single customer. So where do you start? The first thing is to put yourself in the shoes of a hacker and then instrument your APIs to detect and block common attacks along with unknown unknowns for zero-day exploits. Some of these are on the OWASP Security API list, but not all.

Insecure pagination and resource limits

Most APIs provide access to resources that are lists of entities such as /users or /widgets. A client such as a browser would typically filter and paginate through this list to limit the number items returned to a client like so:

First Call: GET /items?skip=0&take=10 
Second Call: GET /items?skip=10&take=10

However, if that entity has any PII or other information, then a hacker could scrape that endpoint to get a dump of all entities in your database. This could be most dangerous if those entities accidently exposed PII or other sensitive information, but could also be dangerous in providing competitors or others with adoption and usage stats for your business or provide scammers with a way to get large email lists. See how Venmo data was scraped

A naive protection mechanism would be to check the take count and throw an error if greater than 100 or 1000. The problem with this is two-fold:

1. For data APIs, legitimate customers may need to fetch and sync a large number of records such as via cron jobs. Artificially small pagination limits can force your API to be very chatty decreasing overall throughput. Max limits are to ensure memory and scalability requirements are met (and prevent certain DDoS attacks), not to guarantee security.
2. This offers zero protection to a hacker that writes a simple script that sleeps a random delay between repeated accesses.

skip = 0
while True:    response = requests.post('https://api.acmeinc.com/widgets?take=10&skip=' + skip),                      headers={'Authorization': 'Bearer' + ' ' + sys.argv[1]})    print("Fetched 10 items")    sleep(randint(100,1000))    skip += 10

How to secure against pagination attacks

To secure against pagination attacks, you should track how many items of a single resource are accessed within a certain time period for each user or API key rather than just at the request level. By tracking API resource access at the user level, you can block a user or API key once they hit a threshold such as “touched 1,000,000 items in a one hour period”. This is dependent on your API use case and can even be dependent on their subscription with you. Like a Captcha, this can slow down the speed that a hacker can exploit your API, like a Captcha if they have to create a new user account manually to create a new API key.

Insecure API key generation

Most APIs are protected by some sort of API key or JWT (JSON Web Token). This provides a natural way to track and protect your API as API security tools can detect abnormal API behavior and block access to an API key automatically. However, hackers will want to outsmart these mechanisms by generating and using a large pool of API keys from a large number of users just like a web hacker would use a large pool of IP addresses to circumvent DDoS protection.

How to secure against API key pools

The easiest way to secure against these types of attacks is by requiring a human to sign up for your service and generate API keys. Bot traffic can be prevented with things like Captcha and 2-Factor Authentication. Unless there is a legitimate business case, new users who sign up for your service should not have the ability to generate API keys programmatically. Instead, only trusted customers should have the ability to generate API keys programmatically. Go one step further and ensure any anomaly detection for abnormal behavior is done at the user and account level, not just for each API key.

Accidental key exposure

APIs are used in a way that increases the probability credentials are leaked:

1. APIs are expected to be accessed over indefinite time periods, which increases the probability that a hacker obtains a valid API key that’s not expired. You save that API key in a server environment variable and forget about it. This is a drastic contrast to a user logging into an interactive website where the session expires after a short duration.
2. The consumer of an API has direct access to the credentials such as when debugging via Postman or CURL. It only takes a single developer to accidently copy/pastes the CURL command containing the API key into a public forum like in GitHub Issues or Stack Overflow.
3. API keys are usually bearer tokens without requiring any other identifying information. APIs cannot leverage things like one-time use tokens or 2-factor authentication.

If a key is exposed due to user error, one may think you as the API provider has any blame. However, security is all about reducing surface area and risk. Treat your customer data as if it’s your own and help them by adding guards that prevent accidental key exposure.

How to prevent accidental key exposure

The easiest way to prevent key exposure is by leveraging two tokens rather than one. A refresh token is stored as an environment variable and can only be used to generate short lived access tokens. Unlike the refresh token, these short lived tokens can access the resources, but are time limited such as in hours or days.

The customer will store the refresh token with other API keys. Then your SDK will generate access tokens on SDK init or when the last access token expires. If a CURL command gets pasted into a GitHub issue, then a hacker would need to use it within hours reducing the attack vector (unless it was the actual refresh token which is low probability)

Exposure to DDoS attacks

APIs open up entirely new business models where customers can access your API platform programmatically. However, this can make DDoS protection tricky. Most DDoS protection is designed to absorb and reject a large number of requests from bad actors during DDoS attacks but still need to let the good ones through. This requires fingerprinting the HTTP requests to check against what looks like bot traffic. This is much harder for API products as all traffic looks like bot traffic and is not coming from a browser where things like cookies are present.

Stopping DDoS attacks

The magical part about APIs is almost every access requires an API Key. If a request doesn’t have an API key, you can automatically reject it which is lightweight on your servers (Ensure authentication is short circuited very early before later middleware like request JSON parsing). So then how do you handle authenticated requests? The easiest is to leverage rate limit counters for each API key such as to handle X requests per minute and reject those above the threshold with a 429 HTTP response. There are a variety of algorithms to do this such as leaky bucket and fixed window counters.

Incorrect server security

APIs are no different than web servers when it comes to good server hygiene. Data can be leaked due to misconfigured SSL certificate or allowing non-HTTPS traffic. For modern applications, there is very little reason to accept non-HTTPS requests, but a customer could mistakenly issue a non HTTP request from their application or CURL exposing the API key. APIs do not have the protection of a browser so things like HSTS or redirect to HTTPS offer no protection.

How to ensure proper SSL

Test your SSL implementation over at Qualys SSL Test or similar tool. You should also block all non-HTTP requests which can be done within your load balancer. You should also remove any HTTP headers scrub any error messages that leak implementation details. If your API is used only by your own apps or can only be accessed server-side, then review Authoritative guide to Cross-Origin Resource Sharing for REST APIs

Incorrect caching headers

APIs provide access to dynamic data that’s scoped to each API key. Any caching implementation should have the ability to scope to an API key to prevent cross-pollution. Even if you don’t cache anything in your infrastructure, you could expose your customers to security holes. If a customer with a proxy server was using multiple API keys such as one for development and one for production, then they could see cross-pollinated data.

#api management #api security #api best practices #api providers #security analytics #api management policies #api access tokens #api access #api security risks #api access keys

Public ASX100 APIs: The Essential List

We’ve conducted some initial research into the public APIs of the ASX100 because we regularly have conversations about what others are doing with their APIs and what best practices look like. Being able to point to good local examples and explain what is happening in Australia is a key part of this conversation.

Method

The method used for this initial research was to obtain a list of the ASX100 (as of 18 September 2020). Then work through each company looking at the following:

1. Whether the company had a public API: this was found by googling “[company name] API” and “[company name] API developer” and “[company name] developer portal”. Sometimes the company’s website was navigated or searched.
2. Some data points about the API were noted, such as the URL of the portal/documentation and the method they used to publish the API (portal, documentation, web page).
3. Observations were recorded that piqued the interest of the researchers (you will find these below).
4. Other notes were made to support future research.
5. You will find a summary of the data in the infographic below.

Data

With regards to how the APIs are shared:

• Documentation is where just the Swagger/OpenAPI docs are hosted somewhere. e.g. https://api.linkgroup.com/member/docs/index
• Portal is where an interactive portal has been developed. e.g. https://dev.telstra.com/
• The web page is where the company has a single page that isn’t interactive explaining their Public APIs. e.g. https://www.bendigoadelaide.com.au/banking-products-api.asp

#api #api-development #api-analytics #apis #api-integration #api-testing #api-security #api-gateway