Month: February 2025

Article originally posted on InfoQ. Visit InfoQ

FerretDB has announced the first release candidate of version 2.0. Now powered by the recently released DocumentDB, FerretDB serves as an open-source alternative to MongoDB, bringing significant performance improvements, enhanced feature compatibility, vector search capabilities, and replication support.

Originally launched as MangoDB three years ago, FerretDB became generally available last year, as previously reported by InfoQ. Peter Farkas, co-founder and CEO of FerretDB, writes:

FerretDB 2.0 represents a leap forward in terms of performance and compatibility. Thanks to changes under the hood, FerretDB is now up to 20x faster for certain workloads, making it as performant as leading alternatives on the market. Users who may have encountered compatibility issues in previous versions will be pleased to find that FerretDB now supports a wider range of applications, allowing more apps to work seamlessly.

Released under the Apache 2.0 license, FerretDB is usually compatible with MongoDB drivers and tools. It is designed as a drop-in replacement for MongoDB 5.0+ for many open-source and early-stage commercial projects that prefer to avoid the SSPL license, a source-available copyleft software license.

FerretDB 2.x is leveraging Microsoft’s DocumentDB PostgreSQL extension. This open-source extension, licensed under MIT, introduces the BSON data type and related operations to PostgreSQL. The solution includes two PostgreSQL extensions: pg_documentdb_core for BSON optimization and pg_documentdb_api for data operations.

According to the FerretDB team, maintaining compatibility between DocumentDB and FerretDB allows users to run document database workloads on Postgres with improved performance and better support for existing applications. Describing the engine behind the vCore-based Azure Cosmos DB for MongoDB, Abinav Rameesh, principal product manager at Azure, explains:

Users looking for a ready-to-use NoSQL database can leverage an existing solution in FerretDB (…) While users can interact with DocumentDB through Postgres, FerretDB 2.0 provides an interface with a document database protocol.

In a LinkedIn comment, Farkas adds:

With Microsoft’s open sourcing of DocumentDB, we are closer than ever to an industry-wide collaboration on creating an open standard for document databases.

In a separate article, Farkas explains why he believes document databases need standardization beyond just being “MongoDB-compatible.” FerretDB provides a list of known differences from MongoDB, noting that while it uses the same protocol error names and codes, the exact error messages may differ in some cases. Although integration with DocumentDB improves performance, it represents a significant shift and introduces regression constraints compared to FerretDB 1.0. Farkas writes:

With the release of FerretDB 2.0, we are now focusing exclusively on supporting PostgreSQL databases utilizing DocumentDB (…) However, for those who rely on earlier versions and backends, FerretDB 1.x remains available on our GitHub repository, and we encourage the community to continue contributing to its development or fork and extend it on their own.

As part of the FerretDB 2.0 launch, FerretDB Cloud is in development. This managed database-as-a-service option will initially be available on AWS and GCP, with support for Microsoft Azure planned for a later date. The high-level road map of the FerreDB project is available on GitHub.

Posted on mongodb google news. Visit mongodb google news

abrdn plc lessened its stake in shares of MongoDB, Inc. (NASDAQ:MDB – Free Report) by 16.7% in the 4th quarter, according to its most recent disclosure with the SEC. The institutional investor owned 34,877 shares of the company’s stock after selling 6,998 shares during the period. abrdn plc’s holdings in MongoDB were worth $8,184,000 at the end of the most recent quarter.

Several other hedge funds and other institutional investors also recently bought and sold shares of MDB. B.O.S.S. Retirement Advisors LLC purchased a new stake in MongoDB during the fourth quarter worth $606,000. Aigen Investment Management LP bought a new position in shares of MongoDB during the 3rd quarter worth approximately $1,045,000. Geode Capital Management LLC boosted its stake in MongoDB by 2.9% in the third quarter. Geode Capital Management LLC now owns 1,230,036 shares of the company’s stock valued at $331,776,000 after acquiring an additional 34,814 shares in the last quarter. B. Metzler seel. Sohn & Co. Holding AG acquired a new stake in MongoDB in the third quarter valued at approximately $4,366,000. Finally, Charles Schwab Investment Management Inc. increased its position in shares of MongoDB by 2.8% during the third quarter. Charles Schwab Investment Management Inc. now owns 278,419 shares of the company’s stock worth $75,271,000 after purchasing an additional 7,575 shares in the last quarter. Institutional investors own 89.29% of the company’s stock.

Insider Transactions at MongoDB

In other news, CAO Thomas Bull sold 169 shares of the firm’s stock in a transaction that occurred on Thursday, January 2nd. The shares were sold at an average price of $234.09, for a total transaction of $39,561.21. Following the completion of the sale, the chief accounting officer now owns 14,899 shares in the company, valued at $3,487,706.91. This trade represents a 1.12 % decrease in their position. The sale was disclosed in a filing with the Securities & Exchange Commission, which is available at this hyperlink. Also, Director Dwight A. Merriman sold 3,000 shares of MongoDB stock in a transaction that occurred on Monday, February 3rd. The shares were sold at an average price of $266.00, for a total transaction of $798,000.00. Following the sale, the director now directly owns 1,113,006 shares in the company, valued at approximately $296,059,596. The trade was a 0.27 % decrease in their position. The disclosure for this sale can be found here. Over the last three months, insiders have sold 42,491 shares of company stock valued at $11,543,480. Corporate insiders own 3.60% of the company’s stock.

MongoDB Trading Down 0.1 %

MongoDB stock traded down $0.23 during midday trading on Friday, reaching $277.87. The company had a trading volume of 1,285,901 shares, compared to its average volume of 1,481,443. MongoDB, Inc. has a 12-month low of $212.74 and a 12-month high of $509.62. The firm’s fifty day simple moving average is $267.08 and its 200-day simple moving average is $270.55.

MongoDB (NASDAQ:MDB – Get Free Report) last issued its quarterly earnings results on Monday, December 9th. The company reported $1.16 earnings per share (EPS) for the quarter, topping analysts’ consensus estimates of $0.68 by $0.48. MongoDB had a negative return on equity of 12.22% and a negative net margin of 10.46%. The firm had revenue of $529.40 million for the quarter, compared to analyst estimates of $497.39 million. During the same period in the previous year, the company posted $0.96 earnings per share. The business’s revenue was up 22.3% on a year-over-year basis. On average, sell-side analysts predict that MongoDB, Inc. will post -1.78 EPS for the current fiscal year.

Wall Street Analyst Weigh In

A number of analysts have recently weighed in on the company. Guggenheim raised MongoDB from a “neutral” rating to a “buy” rating and set a $300.00 price objective for the company in a research report on Monday, January 6th. Tigress Financial lifted their price objective on shares of MongoDB from $400.00 to $430.00 and gave the stock a “buy” rating in a research note on Wednesday, December 18th. Citigroup upped their target price on MongoDB from $400.00 to $430.00 and gave the company a “buy” rating in a research report on Monday, December 16th. Rosenblatt Securities initiated coverage on MongoDB in a research report on Tuesday, December 17th. They set a “buy” rating and a $350.00 price target for the company. Finally, Royal Bank of Canada increased their price objective on MongoDB from $350.00 to $400.00 and gave the company an “outperform” rating in a report on Tuesday, December 10th. Two analysts have rated the stock with a sell rating, four have given a hold rating, twenty-three have issued a buy rating and two have issued a strong buy rating to the company. Based on data from MarketBeat, the company has a consensus rating of “Moderate Buy” and a consensus price target of $361.00.

Get Our Latest Research Report on MongoDB

MongoDB Company Profile

(Free Report)

MongoDB, Inc, together with its subsidiaries, provides general purpose database platform worldwide. The company provides MongoDB Atlas, a hosted multi-cloud database-as-a-service solution; MongoDB Enterprise Advanced, a commercial database server for enterprise customers to run in the cloud, on-premises, or in a hybrid environment; and Community Server, a free-to-download version of its database, which includes the functionality that developers need to get started with MongoDB.

Recommended Stories

Before you consider MongoDB, you’ll want to hear this.

MarketBeat keeps track of Wall Street’s top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis. MarketBeat has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on… and MongoDB wasn’t on the list.

While MongoDB currently has a “Moderate Buy” rating among analysts, top-rated analysts believe these five stocks are better buys.

View The Five Stocks Here

As the AI market heats up, investors who have a vision for artificial intelligence have the potential to see real returns. Learn about the industry as a whole as well as seven companies that are getting work done with the power of AI.

Get This Free Report

Article originally posted on mongodb google news. Visit mongodb google news

Posted on nosqlgooglealerts. Visit nosqlgooglealerts

In today’s digital landscape, scalability is a crucial factor for web applications. As businesses expand, their data grows exponentially, making traditional relational databases less efficient for handling large-scale applications. 

This is where NoSQL databases come into play. The best NoSQL databases for scalable web applications offer high performance, flexibility, and distributed architecture to meet modern development needs.

Unlike traditional relational databases that rely on structured schemas and SQL queries, NoSQL databases provide diverse data storage solutions that accommodate unstructured, semi-structured, and structured data. 

These databases are designed to handle rapid growth, making them ideal for businesses dealing with massive datasets, real-time analytics, and cloud-based applications.

In this comprehensive guide, we will explore the 5 best NoSQL databases for scalable web applications, highlighting their features, use cases, and benefits. By the end of this article, you’ll have a clear understanding of which NoSQL database suits your specific needs.

What Makes a NoSQL Database Ideal for Scalable Web Applications?

Before diving into the best NoSQL databases for scalable web applications, it’s essential to understand what makes them suitable for scalability. NoSQL databases differ from traditional relational databases in several key ways:

High Performance & Speed

• NoSQL databases handle large volumes of data with low latency.
• They use optimized indexing and in-memory storage to enhance query speed.

Flexible Data Modeling

• Unlike relational databases, NoSQL allows dynamic schema changes.
• Supports a variety of data models: document, key-value, column-family, and graph.

Scalability & Horizontal Expansion

• NoSQL databases support horizontal scaling by distributing data across multiple nodes.
• This ensures the system remains highly available even under heavy workloads.

Fault Tolerance & Reliability

• Built-in replication and auto-failover mechanisms ensure high availability.
• Data is distributed across multiple servers to prevent single points of failure.

Now, let’s explore the 5 best NoSQL databases for scalable web applications and how they cater to different business needs.

The 5 Best NoSQL Databases for Scalable Web Applications

Selecting the right NoSQL database is essential for web applications that require high availability, low latency, and seamless scalability. As different NoSQL databases cater to different business needs, understanding their unique strengths is crucial. Below, we explore the best NoSQL databases for scalable web applications, each excelling in specific use cases.

1. MongoDB – Best for General Purpose & High Scalability

MongoDB is one of the most popular NoSQL databases, designed for flexibility and high-performance scalability. It stores data in a JSON-like document format, making it an excellent choice for developers who require fast and efficient data retrieval. Organizations such as eBay, Forbes, and Adobe use MongoDB for its ease of scalability and strong querying capabilities.

Key Features:

• Document-oriented NoSQL database.
• Supports automatic sharding and replication.
• Schema-less design for flexible data modeling.
• Indexing and aggregation framework for fast queries.

Use Cases:

2. Cassandra – Best for High Availability & Distributed Applications

Apache Cassandra is a decentralized, highly available NoSQL database used by companies such as Netflix, Facebook, and Twitter. It is designed to handle massive amounts of data across multiple data centers, ensuring no single point of failure.

Key Features:

• Decentralized peer-to-peer architecture.
• Supports automatic replication across multiple data centers.
• Handles massive read and write operations efficiently.
• No single point of failure.

Use Cases:

• IoT and time-series data storage.
• High-traffic websites.
• Decentralized applications.

3. Redis – Best for Caching & Real-Time Processing

Redis is an in-memory key-value store that provides ultra-low latency and high-speed performance. Used by companies like GitHub, Stack Overflow, and Twitter, Redis is particularly effective for caching, session storage, and real-time analytics.

Key Features:

• In-memory key-value store.
• Ultra-low latency for high-speed applications.
• Supports transactions and pub/sub messaging.
• High-performance caching for API responses.

Use Cases:

• Session management.
• Leaderboards in gaming applications.
• Caching for dynamic web applications.

4. CouchDB – Best for Offline-First & Synchronization Features

CouchDB is an open-source NoSQL database optimized for offline-first applications, synchronization, and mobile solutions. Used by BBC, Credit Suisse, and NHS, CouchDB offers reliable multi-master replication.

Key Features:

• JSON-based document storage.
• Multi-master replication support.
• Provides automatic conflict resolution.
• RESTful HTTP API for seamless integration.

Use Cases:

• Mobile and offline-first applications.
• Distributed collaboration tools.
• Healthcare and financial services.

5. DynamoDB – Best for Fully Managed & Serverless Architectures

Amazon DynamoDB is a cloud-based NoSQL database offering seamless scalability and high availability. Used by Lyft, Airbnb, and Samsung, DynamoDB is fully managed and supports on-demand scaling.

Key Features:

• Fully managed NoSQL database by AWS.
• Auto-scaling and on-demand capacity provisioning.
• Built-in security, backup, and restore features.
• Integration with AWS ecosystem for enhanced performance.

Use Cases:

Takeaways

The best NoSQL databases for scalable web applications provide the flexibility, speed, and fault tolerance necessary to handle modern workloads. Whether you need high availability, low-latency caching, or distributed architecture, there’s a NoSQL database suited for your application.

Understanding your business needs and application requirements will help you choose the best NoSQL database for scalable web applications, ensuring optimal performance and reliability.

Ready to implement the right NoSQL solution? Start experimenting with these databases today and scale your web applications effectively!

Posted on mongodb google news. Visit mongodb google news

Artificial intelligence is the greatest investment opportunity of our lifetime. The time to invest in groundbreaking AI is now, and this stock is a steal!

The whispers are turning into roars.

Artificial intelligence isn’t science fiction anymore.

It’s the revolution reshaping every industry on the planet.

From driverless cars to medical breakthroughs, AI is on the cusp of a global explosion, and savvy investors stand to reap the rewards.

Here’s why this is the prime moment to jump on the AI bandwagon:

Exponential Growth on the Horizon: Forget linear growth – AI is poised for a hockey stick trajectory.

Imagine every sector, from healthcare to finance, infused with superhuman intelligence.

We’re talking disease prediction, hyper-personalized marketing, and automated logistics that streamline everything.

This isn’t a maybe – it’s an inevitability.

Early investors will be the ones positioned to ride the wave of this technological tsunami.

Ground Floor Opportunity: Remember the early days of the internet?

Those who saw the potential of tech giants back then are sitting pretty today.

AI is at a similar inflection point.

We’re not talking about established players – we’re talking about nimble startups with groundbreaking ideas and the potential to become the next Google or Amazon.

This is your chance to get in before the rockets take off!

Disruption is the New Name of the Game: Let’s face it, complacency breeds stagnation.

AI is the ultimate disruptor, and it’s shaking the foundations of traditional industries.

The companies that embrace AI will thrive, while the dinosaurs clinging to outdated methods will be left in the dust.

As an investor, you want to be on the side of the winners, and AI is the winning ticket.

The Talent Pool is Overflowing: The world’s brightest minds are flocking to AI.

From computer scientists to mathematicians, the next generation of innovators is pouring its energy into this field.

This influx of talent guarantees a constant stream of groundbreaking ideas and rapid advancements.

By investing in AI, you’re essentially backing the future.

The future is powered by artificial intelligence, and the time to invest is NOW.

Don’t be a spectator in this technological revolution.

Dive into the AI gold rush and watch your portfolio soar alongside the brightest minds of our generation.

This isn’t just about making money – it’s about being part of the future.

So, buckle up and get ready for the ride of your investment life!

Act Now and Unlock a Potential 10,000% Return: This AI Stock is a Diamond in the Rough (But Our Help is Key!)

The AI revolution is upon us, and savvy investors stand to make a fortune.

But with so many choices, how do you find the hidden gem – the company poised for explosive growth?

That’s where our expertise comes in.

We’ve got the answer, but there’s a twist…

Imagine an AI company so groundbreaking, so far ahead of the curve, that even if its stock price quadrupled today, it would still be considered ridiculously cheap.

That’s the potential you’re looking at. This isn’t just about a decent return – we’re talking about a 10,000% gain over the next decade!

Our research team has identified a hidden gem – an AI company with cutting-edge technology, massive potential, and a current stock price that screams opportunity.

This company boasts the most advanced technology in the AI sector, putting them leagues ahead of competitors.

It’s like having a race car on a go-kart track.

They have a strong possibility of cornering entire markets, becoming the undisputed leader in their field.

Here’s the catch (it’s a good one): To uncover this sleeping giant, you’ll need our exclusive intel.

We want to make sure none of our valued readers miss out on this groundbreaking opportunity!

That’s why we’re slashing the price of our Premium Readership Newsletter by a whopping 70%.

For a ridiculously low price of just $29.99, you can unlock a year’s worth of in-depth investment research and exclusive insights – that’s less than a single restaurant meal!

Here’s why this is a deal you can’t afford to pass up:

• Access to our Detailed Report on this Game-Changing AI Stock: Our in-depth report dives deep into our #1 AI stock’s groundbreaking technology and massive growth potential.

• 11 New Issues of Our Premium Readership Newsletter: You will also receive 11 new issues and at least one new stock pick per month from our monthly newsletter’s portfolio over the next 12 months. These stocks are handpicked by our research director, Dr. Inan Dogan.

• One free upcoming issue of our 70+ page Quarterly Newsletter: A value of $149

• Bonus Reports: Premium access to members-only fund manager video interviews

• Ad-Free Browsing: Enjoy a year of investment research free from distracting banner and pop-up ads, allowing you to focus on uncovering the next big opportunity.

• 30-Day Money-Back Guarantee:  If you’re not absolutely satisfied with our service, we’ll provide a full refund within 30 days, no questions asked.

Space is Limited! Only 1000 spots are available for this exclusive offer. Don’t let this chance slip away – subscribe to our Premium Readership Newsletter today and unlock the potential for a life-changing investment.

Here’s what to do next:

1. Head over to our website and subscribe to our Premium Readership Newsletter for just $29.99.

2. Enjoy a year of ad-free browsing, exclusive access to our in-depth report on the revolutionary AI company, and the upcoming issues of our Premium Readership Newsletter over the next 12 months.

3. Sit back, relax, and know that you’re backed by our ironclad 30-day money-back guarantee.

Don’t miss out on this incredible opportunity! Subscribe now and take control of your AI investment future!

No worries about auto-renewals! Our 30-Day Money-Back Guarantee applies whether you’re joining us for the first time or renewing your subscription a year later!

Article originally posted on mongodb google news. Visit mongodb google news

Posted on mongodb google news. Visit mongodb google news

Over the past six months, MongoDB has been a great trade, beating the S&P 500 by 6.8%. Its stock price has climbed to $278.16, representing a healthy 23.7% increase. This was partly thanks to its solid quarterly results, and the performance may have investors wondering how to approach the situation.

Is now still a good time to buy MDB? Or is this a case of a company fueled by heightened investor enthusiasm? Find out in our full research report, it’s free.

Why Does MongoDB Spark Debate?

Started in 2007 by the team behind Google’s ad platform, DoubleClick, MongoDB offers database-as-a-service that helps companies store large volumes of semi-structured data.

Two Positive Attributes:

1. Billings Surge, Boosting Cash On Hand

Billings is a non-GAAP metric that is often called “cash revenue” because it shows how much money the company has collected from customers in a certain period. This is different from revenue, which must be recognized in pieces over the length of a contract.

MongoDB’s billings punched in at $511.6 million in Q3, and over the last four quarters, its year-on-year growth averaged 26.1%. This performance was fantastic, indicating robust customer demand. The high level of cash collected from customers also enhances liquidity and provides a solid foundation for future investments and growth.

2. Outstanding Retention Sets the Stage for Huge Gains

One of the best parts about the software-as-a-service business model (and a reason why they trade at high valuation multiples) is that customers typically spend more on a company’s products and services over time.

MongoDB’s net revenue retention rate, a key performance metric measuring how much money existing customers from a year ago are spending today, was 120% in Q3. This means MongoDB would’ve grown its revenue by 19.8% even if it didn’t win any new customers over the last 12 months.

MongoDB has a good net retention rate, proving that customers are satisfied with its software and getting more value from it over time, which is always great to see.

One Reason to be Careful:

Long Payback Periods Delay Returns

The customer acquisition cost (CAC) payback period measures the months a company needs to recoup the money spent on acquiring a new customer. This metric helps assess how quickly a business can break even on its sales and marketing investments.

MongoDB’s recent customer acquisition efforts haven’t yielded returns as its CAC payback period was negative this quarter, meaning its sales and marketing investments outpaced its revenue. This inefficiency partly stems from its focus on enterprise clients who require some degree of customization, resulting in long onboarding periods. The complex integrations are a double-edged sword – while MongoDB may not see immediate returns from its sales and marketing investments, it is rewarded with higher switching costs and lifetime value if it can continue meeting its customer’s needs.

Final Judgment

MongoDB has huge potential even though it has some open questions, and with its shares beating the market recently, the stock trades at 9.3× forward price-to-sales (or $278.16 per share). Is now the right time to buy? See for yourself in our comprehensive research report, it’s free.

Stocks We Like Even More Than MongoDB

The Trump trade may have passed, but rates are still dropping and inflation is still cooling. Opportunities are ripe for those ready to act – and we’re here to help you pick them.

Get started by checking out our Top 5 Strong Momentum Stocks for this week. This is a curated list of our High Quality stocks that have generated a market-beating return of 175% over the last five years.

Stocks that made our list in 2019 include now familiar names such as Nvidia (+2,183% between December 2019 and December 2024) as well as under-the-radar businesses like Comfort Systems (+751% five-year return). Find your next big winner with StockStory today for free.

Article originally posted on mongodb google news. Visit mongodb google news

Posted on mongodb google news. Visit mongodb google news

Software is rapidly reducing operating expenses for businesses. Companies bringing it to life have been rewarded with explosive earnings growth, and the upward trend shows no signs of stopping as the industry has posted a 42.2% gain over the past six months, beating the S&P 500 by 25.3 percentage points.

Nevertheless, investors should tread carefully as AI will commoditize many software products, and backing the wrong horse could result in hefty losses. With that said, here is one resilient software stock at the top of our wish list and two we’re steering clear of.

Two Software-as-a-Service Stocks to Sell:

Amplitude (AMPL)

Market Cap: $1.58 billion

Born out of a failed voice recognition startup by founder Spenser Skates, Amplitude (NASDAQ:AMPL) is data analytics software helping companies improve and optimize their digital products.

Why Are We Cautious About AMPL?

1. Average billings growth of 5.5% was subpar, suggesting it’s struggled to push its software and might have to lower prices to stimulate demand
2. Net revenue retention rate of 97.3% shows it has a tough time retaining customers
3. Track record of operating losses stem from its decision to pursue growth instead of profits

Amplitude’s stock price of $12.36 implies a valuation ratio of 4.9x forward price-to-sales. Check out our free in-depth research report to learn more about why AMPL doesn’t pass our bar.

Palo Alto Networks (PANW)

Market Cap: $123 billion

Founded in 2005 by cybersecurity engineer Nir Zuk, Palo Alto Networks (NASDAQ:PANW) makes hardware and software cybersecurity products that protect companies from cyberattacks, breaches, and malware threats.

Why Are We Wary of PANW?

• Products, pricing, or go-to-market strategy may need some adjustments as its 4.2% average billings growth is weak

At $190 per share, Palo Alto Networks trades at 7x forward price-to-sales. Dive into our free research report to see why there are better opportunities than PANW.

One Software-as-a-Service Stock to Watch:

MongoDB (MDB)

Market Cap: $20.71 billion

Started in 2007 by the team behind Google’s ad platform, DoubleClick, MongoDB offers database-as-a-service that helps companies store large volumes of semi-structured data.

Why Are We Fans of MDB?

1. Winning new contracts that can potentially increase in value as its billings growth has averaged 26.1% over the last year
2. Customers use its software daily and increase their spending every year, as seen in its 120% net revenue retention rate
3. Estimated revenue growth of 15.7% for the next 12 months implies its momentum over the last three years will continue

MongoDB is trading at $278.33 per share, or 9.3x forward price-to-sales. Is now a good time to buy? See for yourself in our full research report, it’s free.

Stocks We Like Even More

The Trump trade may have passed, but rates are still dropping and inflation is still cooling. Opportunities are ripe for those ready to act – and we’re here to help you pick them.

Get started by checking out our Top 5 Strong Momentum Stocks for this week. This is a curated list of our High Quality stocks that have generated a market-beating return of 175% over the last five years.

Stocks that made our list in 2019 include now familiar names such as Nvidia (+2,183% between December 2019 and December 2024) as well as under-the-radar businesses like Sterling Infrastructure (+1,096% five-year return). Find your next big winner with StockStory today for free.

Article originally posted on mongodb google news. Visit mongodb google news

Posted on mongodb google news. Visit mongodb google news

Mirae Asset Global Investments Co. Ltd. raised its position in shares of MongoDB, Inc. (NASDAQ:MDB – Free Report) by 287.1% in the fourth quarter, according to the company in its most recent Form 13F filing with the Securities and Exchange Commission. The firm owned 55,736 shares of the company’s stock after acquiring an additional 41,339 shares during the period. Mirae Asset Global Investments Co. Ltd. owned approximately 0.07% of MongoDB worth $13,130,000 at the end of the most recent reporting period.

Other institutional investors and hedge funds also recently added to or reduced their stakes in the company. Nisa Investment Advisors LLC raised its position in MongoDB by 3.8% in the third quarter. Nisa Investment Advisors LLC now owns 1,090 shares of the company’s stock valued at $295,000 after purchasing an additional 40 shares during the period. Hilltop National Bank boosted its stake in shares of MongoDB by 47.2% in the fourth quarter. Hilltop National Bank now owns 131 shares of the company’s stock valued at $30,000 after purchasing an additional 42 shares during the period. Tanager Wealth Management LLP grew its holdings in MongoDB by 4.7% during the 3rd quarter. Tanager Wealth Management LLP now owns 957 shares of the company’s stock valued at $259,000 after purchasing an additional 43 shares in the last quarter. Rakuten Securities Inc. increased its stake in MongoDB by 16.5% during the 3rd quarter. Rakuten Securities Inc. now owns 332 shares of the company’s stock worth $90,000 after buying an additional 47 shares during the period. Finally, Prime Capital Investment Advisors LLC lifted its holdings in MongoDB by 5.2% in the 3rd quarter. Prime Capital Investment Advisors LLC now owns 1,190 shares of the company’s stock worth $322,000 after buying an additional 59 shares in the last quarter. 89.29% of the stock is currently owned by institutional investors and hedge funds.

Wall Street Analyst Weigh In

Several analysts have recently weighed in on the company. Stifel Nicolaus raised their target price on MongoDB from $325.00 to $360.00 and gave the company a “buy” rating in a research note on Monday, December 9th. DA Davidson raised their price objective on MongoDB from $340.00 to $405.00 and gave the company a “buy” rating in a research note on Tuesday, December 10th. Morgan Stanley upped their target price on MongoDB from $340.00 to $350.00 and gave the company an “overweight” rating in a research report on Tuesday, December 10th. Scotiabank reduced their target price on MongoDB from $350.00 to $275.00 and set a “sector perform” rating for the company in a research note on Tuesday, January 21st. Finally, Truist Financial reaffirmed a “buy” rating and issued a $400.00 price target (up from $320.00) on shares of MongoDB in a report on Tuesday, December 10th. Two research analysts have rated the stock with a sell rating, four have given a hold rating, twenty-three have assigned a buy rating and two have assigned a strong buy rating to the stock. According to data from MarketBeat.com, the stock has a consensus rating of “Moderate Buy” and an average price target of $361.00.

Check Out Our Latest Research Report on MDB

Insider Buying and Selling at MongoDB

In related news, CAO Thomas Bull sold 1,000 shares of the firm’s stock in a transaction dated Monday, December 9th. The stock was sold at an average price of $355.92, for a total transaction of $355,920.00. Following the completion of the transaction, the chief accounting officer now owns 15,068 shares of the company’s stock, valued at $5,363,002.56. This trade represents a 6.22 % decrease in their ownership of the stock. The sale was disclosed in a filing with the SEC, which is available through the SEC website. Also, insider Cedric Pech sold 287 shares of the company’s stock in a transaction dated Thursday, January 2nd. The shares were sold at an average price of $234.09, for a total transaction of $67,183.83. Following the completion of the sale, the insider now directly owns 24,390 shares of the company’s stock, valued at $5,709,455.10. The trade was a 1.16 % decrease in their ownership of the stock. The disclosure for this sale can be found here. In the last 90 days, insiders have sold 42,491 shares of company stock valued at $11,543,480. Company insiders own 3.60% of the company’s stock.

MongoDB Trading Down 1.4 %

Shares of NASDAQ MDB opened at $278.10 on Friday. MongoDB, Inc. has a 12-month low of $212.74 and a 12-month high of $509.62. The company has a market cap of $20.71 billion, a PE ratio of -101.50 and a beta of 1.28. The stock has a 50 day simple moving average of $268.90 and a 200-day simple moving average of $270.32.

MongoDB (NASDAQ:MDB – Get Free Report) last announced its quarterly earnings results on Monday, December 9th. The company reported $1.16 earnings per share (EPS) for the quarter, topping analysts’ consensus estimates of $0.68 by $0.48. The company had revenue of $529.40 million during the quarter, compared to analyst estimates of $497.39 million. MongoDB had a negative return on equity of 12.22% and a negative net margin of 10.46%. The firm’s quarterly revenue was up 22.3% on a year-over-year basis. During the same quarter in the previous year, the firm posted $0.96 EPS. Sell-side analysts forecast that MongoDB, Inc. will post -1.78 earnings per share for the current fiscal year.

MongoDB Company Profile

(Free Report)

MongoDB, Inc, together with its subsidiaries, provides general purpose database platform worldwide. The company provides MongoDB Atlas, a hosted multi-cloud database-as-a-service solution; MongoDB Enterprise Advanced, a commercial database server for enterprise customers to run in the cloud, on-premises, or in a hybrid environment; and Community Server, a free-to-download version of its database, which includes the functionality that developers need to get started with MongoDB.

Further Reading

This instant news alert was generated by narrative science technology and financial data from MarketBeat in order to provide readers with the fastest and most accurate reporting. This story was reviewed by MarketBeat’s editorial team prior to publication. Please send any questions or comments about this story to contact@marketbeat.com.

Before you make your next trade, you’ll want to hear this.

MarketBeat keeps track of Wall Street’s top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis.

Our team has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on… and none of the big name stocks were on the list.

They believe these five stocks are the five best companies for investors to buy now…

See The Five Stocks Here

Before you consider MongoDB, you’ll want to hear this.

MarketBeat keeps track of Wall Street’s top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis. MarketBeat has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on… and MongoDB wasn’t on the list.

While MongoDB currently has a “Moderate Buy” rating among analysts, top-rated analysts believe these five stocks are better buys.

View The Five Stocks Here

Wondering where to start (or end) with AI stocks? These 10 simple stocks can help investors build long-term wealth as artificial intelligence continues to grow into the future.

Get This Free Report

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on InfoQ. Visit InfoQ

Transcript

He: I’m Jun. I’m the tech lead of the data platform at Netflix, doing management and workflow automation.

I will first give a brief introduction of the problem space. Then I will give an overview of the architectural design. After that, I will show some use cases and examples. Finally, summarize the talk with key takeaways and future works.

Data at Netflix (Introduction)

Let’s get started by looking at the landscape of the data insights at Netflix. Are you a Netflix subscriber? As you might experience while navigating netflix.com, the personalized show recommendation results is quite good. It many times, not always, gives you the shows that you are interested in. Those are powered by the data insights based on multiple data pipelines and also machine learning workflows. Netflix is a data driven company.

Many decisions at Netflix are entirely driven by the data insights, data practitioners like data engineer, data scientist, machine learning engineer, software engineer, even non-engineer, like a data producer, they all run their data pipelines to get the insights they need. It can be just the color used in the landing page when you visit netflix.com, or it can be the personalized recommendation, or the content producer may decide if they should renew the lease, or terminate the next season of the show based on data insights.

Also, we scan data to get a security issue or anything. Data is used widely. As the business continues to expand to new areas, like from streaming to games to ads to the lives. You may have watched the recent live events, and so demand for data continues to grow. Those new initiatives also bring lots of new requirements, for example, like security requirement, privacy requirement, or latency requirements. They all bring a wider variety of use cases to the platform as well.

While users are working with data, the data practitioners usually face three common problems: data accuracy, data freshness, and cost efficiency. With a huge amount of data, the data accuracy is critically important. Business decisions have to be made based on the high-quality data. Also, with that amount of data, if there are some issues you have to correct data, then it might be expensive and also time consuming, as you have to backfill a lot of data. Data freshness is also very important. Users need to process large datasets quickly to enable fast business decision. Cost efficiency is always important, as we run the business.

Netflix spends $150 million per year just on the compute and the storage. Also, I would like to call out these three problems: general and common problems. No matter how big or small your data is, it might be more impactful if the data size is large. It costs a lot. If we can solve them, this will be a game changer and enable lots of new patterns, and also allows us to rethink about the batch ETL, in analytics domain. There are lots of challenges to solve those problems. One of the important ones is late arriving data. This graph shows a visual example.

For instance, last night, at 10:20 p.m., I watched Netflix. I opened my app, then my phone’s battery was dead. Then the event generated that my device at 10:20 p.m. won’t be able to send it to the server, so it buffered my device. Then I put my iPhone to the charger, and then I went to sleep. This morning, I got up at 8:20, then I opened my phone and started the Netflix app. Then the event at 8:20 a.m., plus the events generated last night, both sent to Netflix server, got processed. Those events generated last night got processed with multiple hours of delay.

Then, that’s caused trouble. The key is that the event time matters a lot to the business, not the processing time. Many times, the streaming or ingesting pipeline uses the processing time so they can quickly ingest data and append data to the staging table, which is partitioned by the processing time. This can greatly simplify the streaming pipeline, which is wonderful. Then it leaves the late batch analytics pipeline to handle the late arriving data.

As the data lands late, the data processed a few hours ago or in the past, becomes incomplete, which then caused the data accuracy issue. We can fix that by reprocessing the data, but given that amount of data, that might be expensive or time consuming. Also, while we deal with those large datasets, we usually have to carefully design the partition schema to feed the business needs. Then the late arriving data might block the downstream pipelines to start, because those pipelines would like to start to process data only when the data is complete, as much as possible. That will reduce the data freshness as well.

To assist data practitioners at Netflix to work with data, solve the problem or derive any solutions, we developed this big data analytics platform as a high-level abstraction to offer the best user experience and high-level abstractions for users to interact with those compute engines. Users use our Maestro workflow orchestrator, which abstract lots of the complexity from users, they don’t deal with Spark directly. They can easily write their jobs and then use the engine they like to process their data.

Eventually, data is saved to the Iceberg table. We observe the users in our platform usually follow those two common patterns to deal with late arriving data. First is called lookback window. In that scenario, the workflow owner or the job owner, usually have a lot of business domain knowledge, so they can tell how long they should look back. If the data is older, then that window likely does not have much business value there, so they can discard.

Then they can, for example, always reprocess the past three days of data every day to insert overrides to the target table every day to bring back the data accuracy after three days. Another approach is that we can just ignore the late arriving data that sometimes works, especially the business decision we have to make. How do we make, at real time or at that moment? If data is not there, then we have to make a decision, and we’ll make a decision so late arriving data doesn’t matter.

Then we got freshness, we got cost efficiency, but we lose the data accuracy. Another well-known pattern called incremental processing can address those problems. Incremental processing is just an approach to process data, but only new or change data. Here we focus on the analytical use cases. To support that, we have to solve two problems. One is how we can capture the change. Secondly, how we can track the state.

In this talk, I’m going to show how we can use Iceberg plus Maestro together to efficiently support the change capturing and also give users a very great experience to integrate with their pipelines. Let’s talk about Iceberg first. Have you heard about Iceberg? It is a high-performance table format for huge analytics tables. This project started about 8 years ago at Netflix. It has now become the top batch project and one of the most popular open table formats. It brings lots of great features, I’ve listed some here. It simplifies a lot of the data management.

For our project, we leverage the Iceberg metadata layer. It provides lots of information to help us build a mechanism to be able to catch the change without actually reading the user data at all. I’m going to talk about it later. Let’s go over some basic Iceberg table concepts first. Iceberg tables are saved in the catalog. This catalog can be pluggable like a Hive catalog, or Glue catalog, or JDBC, or REST catalog. At Netflix, we have our own internal catalog service as well. Then the tables will save the metadata file in the metadata file which has a list of manifest files which map to the snapshots.

Then those manifest files are just files that save lots of additional information related to the data files. It also keeps the reference of the data file as well. Then Iceberg will produce the partitioning value by taking a column value and optionally transform it. Then the Iceberg tracks this relationship. The table partitioning data design is purely based on these relationships, so it no longer depends on the table’s physical layout. That’s a very important property that we can leverage to build that efficient, incremental change capture feature.

Data practitioners at Netflix used Iceberg and created more than 1 million tables there, and developed hundreds of thousands of workflows to read or write or transform the data of those tables. We then needed to orchestrate those workflows, that’s why we invented Maestro. Maestro is a horizontal scalable workflow orchestrator that manages the large-scale data and machine learning workflows. I

t manages that end-to-end whole lifecycle, so give users a serverless experience they just like to own their business code, and then they ship the code to the platform. It offers multiple reusable patterns, like for each, conditional branching, and subworkflow, and so on. Our users also build additional patterns using those usable patterns. It is also designed for expansion and integration with others, like Metaflow: its integration with Maestro. Then Iceberg, we’re going to talk about that.

We initiated the Maestro project about four years ago. The decision to build our own workflow orchestrator, instead of using those popular ones like Airflow, is just because of the challenges we were facing at Netflix, for example, scalability. We needed a horizontal scalable workflow orchestrator.

Also, usability and accessibility. We had the alpha release in 2021, later in 2022 we have the beta release. Then later in 2022 we got GA internally. Then the team spent one year to move hundreds of thousands of workflows from the old system to Maestro. It’s a fully managed migration. Users don’t actually make any line of code change. After that, summer, this year, we made the Maestro code publicly available, so you can try it out.

I would like to show these simple examples, just to give you some sense how users interact with Maestro or write their workflows at Netflix. This is a configuration, like a definition, where users define this and then they also can include some business logic there. The first section, like description section, that the user can put some information, even some on-call instruction there. This supports Markdown syntax. Then when there’s alerts, or there’s something wrong with this workflow, when we send an email, we can include this description in the email body as well. Then a user can say, I want to trigger it to run daily, so we support a cron trigger and also a single trigger as well.

For example, if the upstream table is ready, then please run this workflow, the single trigger supports. Then here you see, in the workflow, users can define parameters. Parameter can reference another parameter during the runtime evaluation. Here, the parameter of my query includes the SQL query trying to do something. Then a user can define, I will run this query in Spark. Then they just simply put this configuration there. They don’t need to worry about which cluster they need to route to, or what’s the memory or settings they need to use? They can always let the platform decide that. They just need to pass as a query.

Then once they have this workflow definition defined, when they save it, they can use our CI tool trying to push it and then run it. During the development, the query may not be perfect, or they may need multiple iterations. You can just simply run the query, use like SELECT 1 or something first, until you are satisfied with your results, you can plug in the production query. Then a user can also use the UI to take a look at what’s wrong or what happens.

Maestro provides a workflow platform for everyone, serving thousands of internal Netflix users, including engineers and the lang engineers. It offers multiple interfaces and also a flexible integration and dynamic workflow engine, and these extensible execution supports. With all these features, Maestro has become very successful at Netflix as the data and the machine learning workflow orchestrators.

Thousands of our users use that, developed hundreds of thousands of workflows there. It runs half a million jobs, and in some busy days, it even runs 2 million jobs per day. Again, we would like to provide a clean and easy to adopt solution for users to be able to do efficient incremental processing with data accuracy, freshness, and cost efficiency.

Architectural Design

Now let’s start the fun part, architectural design. There are two major goals of this design. Firstly, we need to efficiently capture the change. This is important, not only because of efficiency, but also because lots of times, we cannot access user data because of security requirement or the privacy requirement. In those cases, then this requirement of efficiently capturing change without reading user data becomes really important.

Fortunately, Iceberg provides all these supports in the metadata layer, and helps us to achieve this. Second is, get the best user experiences. You can imagine that that many users develop that amount of workflows in our platform, we cannot break them, or we cannot ask them to make changes significantly. We would like to offer the best user experience. The key is to decouple the change capturing from the user business logic.

In that way, the implementation or the support of the incremental processing can be engine or language agnostic. In that case, users can use whatever language they like or compute engine they like to implement their business logic, and leave the incremental processing to be handled by the platform. Maestro provides all the support to develop this interface. First, let’s see how we can efficiently capture change. As I mentioned, Iceberg metadata provides lots of useful information. The snapshots contain information about like, how many change rows or added data files.

Then the metadata file per data file gives us information about the reference to the data file, and also the upper and lower bound of the change of a given column from that data file and so on. All that information can help us build a very efficient way to capture the change, or capture even the range of the specific column. Then, we only need to access those metadata and get a reference of data file, and then using that, we can build a mechanism to track the changes. Then we can capture those changes. It’s zero data copy, and we don’t touch user data.

The change captured will be included in a table, which then becomes an interface to hand to our users to consume. This table is the same table as the original table, with the same schemas like security access, everything. The only difference is that this table only contains the change data. Then the table name can be a parameter passed to the user job, then the user just consumes everything from this table, then they get the change data.

Next, I will use an example to show how this approach works. Here I have this simple table called db1, table 1, at workload time, T1. As I mentioned, there’s only one single snapshot there, and then it has two manifest files which have five data files there. Those five data files actually map to two partitions. Those partitions are virtual, as I mentioned. The data files are immutable in the storage. Then at query time, you’ve got those virtual table partitions. Yes, I highlighted here.

Then next, you see here, at T2, I got a new snapshot, S1, where here, it either has three new data files appended to this table, and they somehow have the late arriving data, and so they go to the partition P0, and P1, P2. P2 is the new partition, but then data goes to P0 and P1. We want to process using like a traditional hierarchy, you process all the data. We have to select data from P0, P1, P2 everything, which means that we actually reprocess those data files again. It’s not efficient, thinking about that, if this is like 40 days or something, a huge amount of data.

Then, instead using the Iceberg features, we can create a table called ibp table 1, that has the same schema as the original Iceberg table. It is indeed an Iceberg table. Then we add a new snapshot which has the manifest file. At this moment, we don’t create or copy those data files to create a new data file. Instead, we can read the snapshots information from S1 to get a reference of those data files. Then we just simply add reference to the manifest file of the S2 in the new table.

In this way, we can actually reference them without copying the data, it’s zero data copy, and then when a user queries, they SELECT * from this table, they will get three partitions, P0, P1, P2 as well. In this new virtual partitioning, they don’t have those old data files at all, which means that they will not reprocess data. Also, after a while, the platform is responsible for deleting this temporary table after the ETL finishes.

There are some other alternatives to achieve a similar goal, to enable this incremental processing, using Iceberg, like Iceberg with Flink, with Spark Structure Streaming, and so on. We didn’t go with those approaches, mainly just because they are coupled with the engine tightly. This not only requires users to interact with that library using their API or sign implementation client, and adding to their business logic, also requires users sometimes to rewrite their code.

For example, if they use Trino. If they use other engines, they have to rewrite it, use the engine that is supporting incremental processing. I think the underlying implementation, all those implementations are similar. I show some code here. Basically, we load the table and then use the Iceberg APIs, get the snapshots between the two snapshots ID, and with some filters, say we only care about the append publishing information. Then we create an empty table with the same schema as the original one.

Then we scan through the snapshots table, add data files one by one to this new ICDC table. Have that committed. Of course, this code is simple and it’s not like a full production, but it demonstrates the idea clearly. You might consider a similar approach in your solution if applicable.

This new capability enables many patterns. Here are three emerging ones that we observed or discovered at Netflix. Firstly, we are going to talk about incrementally processing the change data and directly append data to the target table. As we have shown, the change data will include reference to the real data file. This table will have reference to the real data files. Then in the ETL workflows, you simply just consume the ICDC tables, SELECT * from it, and then they can get the data file and then append to the target table. It does not need to reprocess the whole provision of P0, P1, P2, and they will bring lots of savings.

Also, in the future, this might be supported by those SQL extensions or something. You can SELECT * between these two steps here or something. We actually are working on this. A second pattern is that we can use this capture change data as a low-level filter. Many times, especially in the analytics domain, the change data itself does not actually give us the full datasets to process, just for example, if I want to get the watch time for all the Netflix users.

In that case, the change data from the past time window only tells me the users that recently watched Netflix but it does not give me total time. However, this change data itself, you will just take a look at the change data and select a unique user ID from it. We are going to get a small set so the user ID, those users at least watching Netflix in that window, with that, while we’re doing the processing, we can use that as a filter. We can join the original table on this table by these user IDs that we find, we can quickly prompt the datasets to process the transform to be able to get the sum easily. Here I show an example.

By looking at the change data, we found that a circle and diamond actually are the keys there in the change data. Then ETL can load the table, quickly filter and find those kind of diamond and circle data points, and then doing the aggregation and save the overwrites to the target table. In this example, I use a simple sum. You might think about some other incremental way to do the sum, but know that this business logic can be anything. It can be very complicated. You may not be able to do that kind of incremental summation or something. This is just for demo.

The third pattern is actually the captured change range parameter, that’s a little bit different from the change data capture. We even don’t need to create a table, instead we just need to get the upper and lower bound of some specific change columns from the source tables. This is very common, especially if you are going to join multiple tables. That is a common pattern in the batch world, where you join two tables, and then you have to find the common range that you need to select all the data between them to be able to run the join, or run any complicated processing.

In that case, we can read the Iceberg metadata to quickly get the range, like min or max from all the tables for a given column. Then we can tell, this is mean. The mean is P1, and max is P2, and then I can load the table, the partition P1, P2, and the partition P1, P2 during the ETL workflow and then during the processing there. This pattern is very useful in the analytics batch workflows, because it’s very common to join many tables, not just two.

Are we done? We are not done yet. Onboarding cost is another major concern. Thousands of our users develop hundreds of thousands of workflows in Maestro, and then we cannot break those workflows or users, or ask thousands of users to rewrite their pipeline. Also, many times, even a user pipeline is not independent or completely independent. Many times, they are dependent on each other. You have multi-stage pipelines.

In those cases, they might have some stage is enabling incremental processing, but some other stages are not enabling incremental processing. We have to support a mix of those pipelines as well. Then those actual costs, like development, operational maintenance, might completely offset the benefit of the incremental processing. Some of the feedback we heard from a user is that, I don’t have benefits to rewrite. Can it just magically work with little changes? The answer is yes. To address those concerns, we provide other interfaces, in addition to table interfaces. There are two new Maestro step types.

One is called IpCapture step type, which encapsulates all the business logic from the platform to be able to capture the change of the table. Then the IpCommit step, which can commit checkpoint based on the IpCapture step information. With this design, user workflows can simply onboard to Maestro IPS support by just adding one job of IpCapture from, and one job of IpCommit after user jobs. As I mentioned, we are going to have the ICDC table include all the changes, and then that table name will be passed as the parameter to the user jobs.

That’s a great interface for users to be able to simply just consume that table. Most user business logic will be exactly same. They don’t need to worry about how to capture the change at all. In case you really need to maintain or upgrade or fix bugs in our change capturing logic, users won’t get impacted. They just need to rerun their job. Everything will work. They don’t need to add any library dependency to get incremental processing support. Also, multi-stage pipelines can work as well, because this is just like a typical, standardized Maestro workflow. All the workflows can work using the Maestro features.

With those interfaces, all existing workflows can work together seamlessly, and so a user can use the best way to implement their business logic, or can use the engine they like. Also, Maestro step is configurable, as I just showed you, with a very low code solution there too. I would show a complicated example here just to demonstrate how powerful and simple that approach is. Here’s a complicated workflow example that our user developed trying to auto remediate the issues.

Many times, in your ETL pipeline, it may fail because of some small problems here, there, and then the on-call will wake up and then rerun some script to fix the data problem or something, and then just kick off the restart again. Then, in this auto remediation approach, you’re just doing the same thing in this type of flow that either cause this typical ETL pipeline plus auditing process. Then, they can tell Maestro, if this subworkflow failed, please don’t fail the workflow and page me, please just ignore that first and then run a recovery job.

This check status job will see if actually subworkflow failed or not. It either goes to this recovery step and run the recovery job, and then run this workflow again to see if it succeeds. If not, then page a user. This is how a user defined that using Maestro. They can go and define this workflow, with this special flag saying, IGNORE_FAILURE. Then they define the data information, which I mentioned, like if else condition here, trying to route the workflow to different paths based on the status of that subworkflow job.

To enable incremental processing for this complicated pattern is very simple. You either just need an IpCapture step at the beginning and IpCommit step at the end, that’s it. They don’t need to actually modify lots of their code. Then you can see, here is the new pipeline. This is subworkflow which points to the workflow that I just showed. It’s an auto remediation pipeline. Then either pass the source table, which has now become the ICDC table name to the subworkflow, where it also can pass a query as well.

Then it adds this IpCapture step trying to capture changes. You just have this ICDC mode, and then the table is membership_table. Then, you only care to append only in snapshot. Then in the commit step, users just need to tell us what’s the step ID of the IpCapture. That’s it. IPS can efficiently capture the incremental change and handle the late arriving data. With those clean interfaces, the solution is compatible with the existing user experience with really low onboarding cost, as I just showed.

Use Cases and Examples

Next, let’s walk over a few use cases and examples together. This is a two-stage ETL pipeline with two Maestro workflows and three tables. This is a playback table which ingests the events from the streaming pipeline. Then the workflow owner decided, I have to take a look back at two weeks of data, because there’s a lot of events that come late. Then in their daily pipeline, they aggregate this table and then save it to target table by changing the partition key from processing to the event time.

Then they have to do this every day, and rewrite the past 14 days of data. This pipeline can take quite a while, so they have to run daily. They cannot run hourly. Then for this table, consumers, we may have hundreds of workflows consumed from this table. They build aggregation pipelines to power their business use cases. Then here in this case, just doing data aggregation, also have to consume the data from the past 14 days as well. You can see it’s time consuming, plus it’s also fragile.

Many times, if there’s a traffic pattern change, suddenly huge events land, or there’s a business logic change, users have to adjust this lookback window, which then will affect all the downstream workflows and everyone as well. That’s another great pattern. Then, let’s see how we can make that to be IPS enabled. Here’s the one that’s using the pattern 1. You can see that we can easily have an ICDC table to hold the change data, and then let this pipeline consume from ICDC table instead of from the original table. Then we can easily merge into the target table.

Also, because this runs so fast, we can be doing this hourly instead of daily. This shows some changes in the write job. I just use SQL as a demonstration. There’s also Scala or some other logic. You can see that instead of insert overwrite we can use merge into. Then just change from the ICDC table, and then add the dedupe logic. Also, you can use insert into if you like, or if there is a dedupe logic or other workflows running, trying to dedupe the data.

For the stage two, we use the pattern 2, where we can have the ICDC table hold the change data, then join with the IPS table, join with the original table, and then doing this aggregation again and merging to target table. Same thing, user can run super-fast. Then we can update the workflows cadence from daily to hourly as well. This shows the change in SQL.

Basically, same thing, merge into, and then we join with the ICDC table on those aggregation group_by_keys. Then we add the dedupe logic as well. After changing this two-stage pipeline from the original way to this new IPS enabled approach, we see huge savings. You see that the lookback window is about 14 days. Also, the late arriving data is sparse. It’s not always that it has lots of late arriving data. Actually, the cost of the new pipeline is less, or only 10% of the original pipeline. Also, it improved the data freshness as we can now run hourly.

The next example shows the multi-table cases using pattern 3, where we have like three staging tables, row tables which have the events from the streaming pipeline. The first one and the last one has late arriving data, and the middle one does not. Middle one can be a normal pipeline, and this first and the third pipeline will be the one with the IPS enabled. Then they produce this hourly table, 1, 2, 3.

Then this final pipeline will be based on the range captured, here it is min of 3 and max is 6, so that you are going to load the data from all three tables from hour 3 to hour 6 together. Then doing the complicated join operations to produce data to the target table. I hope this example demonstrates how simple and powerful it is for users working with Maestro using IPS.

Key Takeaways

IPS enables new patterns, and also let us rethink about batch ETL, which is sometimes considered to be replaced by the streaming pipeline, which also have some other troubles as well. Actually, the batch ETL is still very popular and widely used, especially in the analytics domain. With this IPS support, we can power the batch ETL to have the data accuracy, data freshness, and cost efficiency, all of these. That can fill lots of gaps and enable lots of use cases, even sometimes that we may not need to move to streaming and can stay with the batch.

In the talk, we have shown how we can use Iceberg metadata to be able to efficiently capture change, and then we also have shown the power of decoupling that can help to address concerns differently, and then let the user have minimum changes. Also, with the clean interface that Maestro provides, we provide a great user experience, and also with minimum effort, users can adopt this new approach. Also, hope you can leverage some of those patterns that we discovered in your work, if applicable.

Future Improvements

What’s next? We are going to move some of the implementation from our own side to the Iceberg, like SQL extension to be able to create a view instead of a table, to reduce the maintenance cost, so we don’t need to create a table. Also, we are going to add support for other types of snapshots beyond the just append. Also, we are working with the Iceberg community to add a cookbook to show how to do the change capturing. We are working on the auto cascading data backfill features using the IPS.

Questions and Answers

Participant: I have a question regarding the join operation that you mentioned for incremental processing. If let’s say I’m doing 7 days or 14 days of join, even for the incremental part, you will still need to load the data on the left side and right-hand side, both for 14 days, because the incremental data might be joining with the historical data in the past. How would you be able to achieve 10x of the cost reduction in that case?

He: Data reduction is based on the fact that if you guesstimate the range, you need to do the join. It may not be accurate, or many times you have to be conservative to handle the worst scenario. You actually have a really long window, then you join that. The cost efficiency we get here is mainly for pattern 1 and 2, but not for the data capture range cases. In the capture range cases, it’s more like we can achieve the optimal data accuracy, as we know exactly what’s the change. The cost efficiency gain from there may not be that huge, like 10x.

See more presentations with transcripts

Posted on mongodb google news. Visit mongodb google news

Jennison Associates LLC lessened its stake in shares of MongoDB, Inc. (NASDAQ:MDB – Free Report) by 99.4% in the 4th quarter, according to its most recent disclosure with the SEC. The firm owned 18,590 shares of the company’s stock after selling 3,083,434 shares during the period. Jennison Associates LLC’s holdings in MongoDB were worth $4,328,000 as of its most recent SEC filing.

Several other hedge funds have also bought and sold shares of the stock. GAMMA Investing LLC lifted its position in shares of MongoDB by 178.8% in the third quarter. GAMMA Investing LLC now owns 145 shares of the company’s stock valued at $39,000 after acquiring an additional 93 shares in the last quarter. Fulton Bank N.A. lifted its position in shares of MongoDB by 5.6% in the third quarter. Fulton Bank N.A. now owns 1,199 shares of the company’s stock valued at $324,000 after acquiring an additional 64 shares in the last quarter. CWM LLC lifted its position in shares of MongoDB by 5.0% in the third quarter. CWM LLC now owns 3,269 shares of the company’s stock valued at $884,000 after acquiring an additional 156 shares in the last quarter. Livforsakringsbolaget Skandia Omsesidigt lifted its position in shares of MongoDB by 253.2% in the third quarter. Livforsakringsbolaget Skandia Omsesidigt now owns 558 shares of the company’s stock valued at $151,000 after acquiring an additional 400 shares in the last quarter. Finally, CHICAGO TRUST Co NA purchased a new stake in shares of MongoDB in the third quarter worth about $255,000. Institutional investors own 89.29% of the company’s stock.

Insider Activity at MongoDB

In other MongoDB news, CAO Thomas Bull sold 1,000 shares of the company’s stock in a transaction on Monday, December 9th. The stock was sold at an average price of $355.92, for a total transaction of $355,920.00. Following the transaction, the chief accounting officer now directly owns 15,068 shares in the company, valued at $5,363,002.56. This represents a 6.22 % decrease in their ownership of the stock. The transaction was disclosed in a document filed with the Securities & Exchange Commission, which is available at the SEC website. Also, insider Cedric Pech sold 287 shares of the company’s stock in a transaction on Thursday, January 2nd. The stock was sold at an average price of $234.09, for a total value of $67,183.83. Following the transaction, the insider now owns 24,390 shares in the company, valued at $5,709,455.10. This trade represents a 1.16 % decrease in their ownership of the stock. The disclosure for this sale can be found here. Over the last three months, insiders have sold 42,491 shares of company stock worth $11,543,480. Insiders own 3.60% of the company’s stock.

MongoDB Trading Down 1.4 %

NASDAQ:MDB opened at $278.10 on Friday. The firm has a market cap of $20.71 billion, a P/E ratio of -101.50 and a beta of 1.28. The stock has a 50 day moving average of $268.90 and a two-hundred day moving average of $270.32. MongoDB, Inc. has a 12 month low of $212.74 and a 12 month high of $509.62.

MongoDB (NASDAQ:MDB – Get Free Report) last issued its earnings results on Monday, December 9th. The company reported $1.16 earnings per share for the quarter, topping the consensus estimate of $0.68 by $0.48. MongoDB had a negative return on equity of 12.22% and a negative net margin of 10.46%. The business had revenue of $529.40 million for the quarter, compared to analysts’ expectations of $497.39 million. During the same period in the previous year, the firm posted $0.96 EPS. The business’s revenue for the quarter was up 22.3% compared to the same quarter last year. On average, research analysts forecast that MongoDB, Inc. will post -1.78 earnings per share for the current year.

Analyst Upgrades and Downgrades

MDB has been the topic of a number of research analyst reports. Wells Fargo & Company boosted their target price on MongoDB from $350.00 to $425.00 and gave the stock an “overweight” rating in a research report on Tuesday, December 10th. Wedbush upgraded MongoDB to a “strong-buy” rating in a research report on Thursday, October 17th. Scotiabank decreased their target price on MongoDB from $350.00 to $275.00 and set a “sector perform” rating on the stock in a research report on Tuesday, January 21st. JMP Securities restated a “market outperform” rating and set a $380.00 price objective on shares of MongoDB in a research report on Wednesday, December 11th. Finally, DA Davidson upped their price objective on MongoDB from $340.00 to $405.00 and gave the company a “buy” rating in a research report on Tuesday, December 10th. Two equities research analysts have rated the stock with a sell rating, four have assigned a hold rating, twenty-three have given a buy rating and two have given a strong buy rating to the stock. Based on data from MarketBeat, the stock has an average rating of “Moderate Buy” and a consensus target price of $361.00.

View Our Latest Analysis on MongoDB

MongoDB Profile

(Free Report)

MongoDB, Inc, together with its subsidiaries, provides general purpose database platform worldwide. The company provides MongoDB Atlas, a hosted multi-cloud database-as-a-service solution; MongoDB Enterprise Advanced, a commercial database server for enterprise customers to run in the cloud, on-premises, or in a hybrid environment; and Community Server, a free-to-download version of its database, which includes the functionality that developers need to get started with MongoDB.

Further Reading

Want to see what other hedge funds are holding MDB? Visit HoldingsChannel.com to get the latest 13F filings and insider trades for MongoDB, Inc. (NASDAQ:MDB – Free Report).

This instant news alert was generated by narrative science technology and financial data from MarketBeat in order to provide readers with the fastest and most accurate reporting. This story was reviewed by MarketBeat’s editorial team prior to publication. Please send any questions or comments about this story to contact@marketbeat.com.

Before you make your next trade, you’ll want to hear this.

MarketBeat keeps track of Wall Street’s top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis.

Our team has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on… and none of the big name stocks were on the list.

They believe these five stocks are the five best companies for investors to buy now…

See The Five Stocks Here

Before you consider MongoDB, you’ll want to hear this.

MarketBeat keeps track of Wall Street’s top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis. MarketBeat has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on… and MongoDB wasn’t on the list.

While MongoDB currently has a “Moderate Buy” rating among analysts, top-rated analysts believe these five stocks are better buys.

View The Five Stocks Here

MarketBeat just released its list of 10 cheap stocks that have been overlooked by the market and may be seriously undervalued. Enter your email address and below to see which companies made the list.

Get This Free Report

Article originally posted on mongodb google news. Visit mongodb google news

Article originally posted on InfoQ. Visit InfoQ

A consortium of 20 European research institutions, companies, and EuroHPC centers has launched OpenEuroLLM, an initiative to develop open-source, multilingual large language models (LLMs). Coordinated by Jan Hajič (Charles University, Czechia) and co-led by Peter Sarlin (AMD Silo AI, Finland), the project aims to provide transparent and compliant AI models for commercial and public sector applications.

The project seeks to align with Europe’s regulatory framework while ensuring that AI development remains accessible and adaptable to various needs. By collaborating with organizations such as LAION, OpenML, and open-sci, OpenEuroLLM plans to release models that support linguistic diversity and can be fine-tuned for specific industry and government use cases.

While the project emphasizes openness and accessibility, some experts have questioned its feasibility. Alek Tarkowski, co-founder of Open Future Foundation, pointed out that the 56 million EUR budget, which was not mentioned in the official announcement, raises concerns about whether a consortium of 20 institutions can effectively build competitive foundation models.

Similarly, Daniel Khachab, co-founder and CEO of Choco, criticized the initiative, stating:

20 companies building something together funded by the government is a recipe for failure. No accountability, leadership, or upside. The EU should rather deregulate and put the €56m in top-notch education.

A key aspect of OpenEuroLLM is its commitment to open-source principles, but the extent of this openness remains debatable. The project describes its models as “truly open” meaning not only open weights but also open datasets, training and testing code, and evaluation metrics. However, Alek Tarkowski added:

None of the model builders in the consortium have released models that meet these ambitious standards, and it is uncertain whether a foundation model can be built on open data alone.

The project’s reference to “compliant open-source models” also raises questions. While the AI Act defines open-source AI in terms of open-weight models, OpenEuroLLM suggests a broader approach. Whether it can meet these goals while maintaining technical competitiveness remains unclear.

OpenEuroLLM has been awarded the STEP (Strategic Technologies for Europe Platform) seal and is funded by the European Commission under the Digital Europe Programme. The consortium begins its work on February 1st, 2025. The project’s success will depend on whether it can effectively coordinate its research efforts and deliver models that balance openness, regulatory compliance, and technological performance.