[{"data":1,"prerenderedAt":1328},["ShallowReactive",2],{"active-banner":3,"navbar-featured-partner-blog":24,"blog-\u002Fblog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud":306,"navbar-pricing-featured":516,"blog-authors-\u002Fblog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud":1293,"related-\u002Fblog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud":1309},{"id":4,"title":5,"date":6,"dismissible":7,"extension":8,"link":9,"link2":10,"linkText":11,"linkText2":12,"meta":13,"stem":21,"variant":22,"__hash__":23},"banners\u002Fbanners\u002Flakestream-ufk-launch.md","StreamNative Introduces Lakestream Architecture and Launches Native Kafka Service","2026-04-07",true,"md","\u002Fblog\u002Ffrom-streams-to-lakestreams","https:\u002F\u002Fconsole.streamnative.cloud\u002Fsignup?from=banner_lakestream-launch","Read Announcement","Sign Up Now",{"body":14},{"type":15,"value":16,"toc":17},"minimark",[],{"title":18,"searchDepth":19,"depth":19,"links":20},"",2,[],"banners\u002Flakestream-ufk-launch","default","zRueBGutATZB0ZnFFHwaEV7F0Di4tnZUHhgOiI4cu6k",{"id":25,"title":26,"authors":27,"body":29,"category":289,"createdAt":290,"date":291,"description":292,"extension":8,"featured":7,"image":293,"isDraft":294,"link":290,"meta":295,"navigation":7,"order":296,"path":297,"readingTime":298,"relatedResources":290,"seo":299,"stem":300,"tags":301,"__hash__":305},"blogs\u002Fblog\u002Fstreamnative-recognized-in-the-forrester-wave-streaming-data-platforms-2025.md","StreamNative Recognized as a Contender in The Forrester Wave™: Streaming Data Platforms, Q4 2025",[28],"David Kjerrumgaard",{"type":15,"value":30,"toc":276},[31,39,47,51,67,73,78,81,87,102,109,115,118,124,127,134,140,143,146,157,163,169,172,175,178,184,191,194,197,204,207,210,224,229,233,237,241,245,249,251,268,270],[32,33,35],"h3",{"id":34},"receives-highest-possible-scores-in-both-the-messaging-and-resource-optimization-criteria",[36,37,38],"em",{},"Receives Highest Possible Scores in BOTH the Messaging and Resource Optimization Criteria",[40,41,43],"h2",{"id":42},"introduction",[44,45,46],"strong",{},"Introduction",[48,49,50],"p",{},"Real-time data has become the backbone of modern innovation. As artificial intelligence (AI) and digital services demand instantaneous insights, organizations are realizing that streaming data is no longer optional – it's essential for delivering timely, context-rich experiences. StreamNative's data streaming platform is built precisely for this reality, ensuring data is immediate, reliable, and ready to power critical applications.",[48,52,53,54,63,64],{},"Today, we're excited to announce that Forrester Research has named StreamNative as a Contender in its evaluation, ",[55,56,58],"a",{"href":57},"\u002Freports\u002Frecognized-in-the-forrester-wave-tm-streaming-data-platforms-q4-2025",[36,59,60],{},[44,61,62],{},"The Forrester Wave™: Streaming Data Platforms, Q4 2025",". This report evaluated 15 top streaming data platform providers, and we're proud to share that ",[44,65,66],{},"StreamNative received the highest scores possible—5 out of 5—in both the Messaging and Resource Optimization criteria.",[48,68,69,70],{},"***Forrester's Take: ***",[36,71,72],{},"\"StreamNative is a good fit for enterprises that want an Apache Pulsar implementation that is also compatible with Kafka APIs.\"",[48,74,75],{},[36,76,77],{},"— The Forrester Wave™: Streaming Data Platforms, Q4 2025",[48,79,80],{},"Being recognized in the Forrester Wave is a proud milestone, and for us, it highlights how far StreamNative has come in enabling enterprises to unlock the power of real-time data. In the sections below, we'll dive into what we believe sets StreamNative apart—from our modern architecture and cloud-native design to our open-source foundation and real-time use cases—and how we see these strengths aligning with Forrester's findings.",[40,82,84],{"id":83},"trusted-by-industry-leaders",[44,85,86],{},"Trusted by Industry Leaders",[48,88,89,90,93,94,97,98,101],{},"Companies across industries are already leveraging StreamNative to drive real-time outcomes. Global enterprises like ",[44,91,92],{},"Cisco"," rely on StreamNative to handle massive IoT telemetry, supporting 245 million+ connected devices. Martech leaders such as ",[44,95,96],{},"Iterable"," process billions of events per day with StreamNative for hyper-personalized customer engagement. And in financial services, ",[44,99,100],{},"FICO"," trusts StreamNative to power its real-time fraud detection and analytics pipelines with a secure, scalable streaming backbone.",[48,103,104,105,108],{},"The Forrester report notes that, “",[36,106,107],{},"Customers appreciate the lower infrastructure costs that result from StreamNative’s cost-efficient, Kafka-compatible architecture. Customers note excellent support responsiveness…","”",[40,110,112],{"id":111},"modern-cloud-native-architecture-built-for-scale",[44,113,114],{},"Modern, Cloud-Native Architecture Built for Scale",[48,116,117],{},"From day one, StreamNative was designed with a modern architecture to meet the demanding scale and flexibility requirements of real-time data. Unlike legacy streaming systems that often rely on tightly coupled storage and compute, StreamNative's platform takes a cloud-native approach: it decouples these layers to enable elastic scalability and efficient resource utilization across any environment. The core is powered by Apache Pulsar—a distributed messaging and streaming engine—enhanced with multi-protocol support (including native Apache Kafka API compatibility) to unify diverse data streams under one roof. This means organizations can consolidate siloed messaging systems and handle both high-volume event streams and traditional message queues on a single platform, without sacrificing performance or reliability.",[48,119,120,121,108],{},"Forrester's evaluation described that “",[36,122,123],{},"StreamNative aims to provide a high-performance, multi-protocol streaming data platform: It uses Apache Pulsar with Kafka API compatibility to deliver cost-efficient, real-time applications for enterprises. It appeals to organizations that want a flexible, low-cost streaming solution, due to its focus on scalability and resource optimization, while its investments in Pulsar’s open-source ecosystem and performance optimization make it the primary platform for enterprises wishing to implement Pulsar.",[48,125,126],{},"Our cloud-first, leaderless architecture (with no single broker bottlenecks) and tiered storage model were built to maximize throughput and cost-efficiency for real-time workloads. By separating compute from storage and leveraging distributed object storage, StreamNative can retain huge volumes of event data indefinitely while keeping compute costs in check—effectively providing a flexible, low-cost streaming solution.",[48,128,129,130,133],{},"This modern design not only delivers high performance, but also ensures fault tolerance and geo-distribution out of the box, so enterprises can trust their streaming data is always available and durable. As Forrester’s evaluation noted, StreamNative ",[36,131,132],{},"\"excels at messaging and resource optimization\" and “Its platform supports use cases like real-time analytics and event-driven architectures with robust scalability.","” Our architecture provides the strong foundation that today's real-time applications demand, from ultra-fast data ingestion to seamless scale-out across hybrid and multi-cloud environments.",[40,135,137],{"id":136},"open-source-foundation-and-pulsar-expertise",[44,138,139],{},"Open Source Foundation and Pulsar Expertise",[48,141,142],{},"StreamNative's DNA is rooted in open source innovation. Our founders are the original creators of Apache Pulsar, and we've built our platform with the same open principles: freedom, flexibility, and community-driven innovation. For developers and data teams, this means adopting StreamNative comes with no proprietary lock-in—instead, you get a platform built on open standards and a thriving ecosystem. We offer broad API compatibility (Pulsar, Kafka, JMS, MQTT, and more) so that teams can work with familiar interfaces and integrate StreamNative into existing systems with ease.",[48,144,145],{},"StreamNative is the primary commercial contributor to the Apache Pulsar project and its surrounding ecosystem. We invest heavily in Pulsar's ongoing improvements our investments in Pulsar's open-source ecosystem and performance optimization bolster StreamNative's value. We also foster a vibrant community through initiatives like the Data Streaming Summit and free training resources.",[48,147,148,149,152,153,156],{},"Forrester's assessment noted that StreamNative’s “",[36,150,151],{},"events-driven agents, extensibility, and performance architecture are solid,","” and we're continuing to build on that foundation. ",[44,154,155],{},"We're actively investing in expanding our tooling for observability, governance, schema management, and developer productivity","—areas we recognize as critical for enterprise adoption and where we're committed to accelerating our roadmap.",[48,158,159,160],{},"Being open also means embracing an open ecosystem of technologies. StreamNative actively integrates with the tools and platforms that matter most to our users. We partner with industry leaders like Snowflake, Databricks, Google, and Ververica to ensure our streaming platform works seamlessly with data warehouses, lakehouse storage, and stream processing frameworks. Forrester’s evaluation observed that StreamNative’s ",[36,161,162],{},"\"investments in Pulsar’s open-source ecosystem and performance optimization make it the primary platform for enterprises wishing to implement Pulsar.\"",[40,164,166],{"id":165},"powering-real-time-use-cases-across-industries",[44,167,168],{},"Powering Real-Time Use Cases Across Industries",[48,170,171],{},"One of the greatest validations of StreamNative's approach is the success our customers are achieving with real-time data. StreamNative's platform is versatile and use-case agnostic—if an application demands high-volume, low-latency data movement, we can power it. This flexibility is why our customer base spans industries from finance and IoT to major automobile manufacturers and online gaming. The common thread is that these organizations need to process and react to data in milliseconds, and StreamNative is delivering the capabilities to make that possible.",[48,173,174],{},"Cisco uses StreamNative to underpin an IoT telemetry system of colossal scale, connecting hundreds of millions of devices and thousands of enterprise clients with real-time data streams. The platform's multi-tenant design and proven reliability allow Cisco to offer its customers a live feed of device data with unwavering confidence. In the financial sector, FICO has built streaming pipelines on StreamNative to detect fraud as transactions happen and to monitor systems in real time. With StreamNative's strong guarantees around message durability and ordering, FICO can catch anomalies or suspicious patterns within seconds. And in digital customer engagement, Iterable relies on StreamNative to process billions of events every day—clicks, views, purchases—so that marketers can trigger personalized campaigns instantly based on user behavior.",[48,176,177],{},"Our customers uniformly deal with mission-critical data streams, where downtime or delays are unacceptable. StreamNative's fault-tolerant, scalable infrastructure has proven equal to the task, handling scenarios like bursting to millions of events per second or seamlessly spanning multiple cloud regions. Forrester's report recognized StreamNative for supporting event-driven architectures with robust scalability—which for us is a reflection of our platform's ability to meet the most demanding enterprise requirements.",[40,179,181],{"id":180},"continuing-to-innovate-ursa-orca-and-the-road-ahead",[44,182,183],{},"Continuing to Innovate: Ursa, Orca, and the Road Ahead",[48,185,186,187,190],{},"While we are thrilled to be recognized in Forrester's Streaming Data Platforms Wave, we view this as just the beginning. StreamNative's vision has always been bold: to ",[44,188,189],{},"provide a unified platform that not only handles today's streaming needs but also anticipates the emerging requirements of tomorrow",".",[48,192,193],{},"One key area of focus is the convergence of streaming data with advanced analytics and AI. As Forrester points out in the report, technology leaders should look for platforms that natively integrate messaging, stream processing, and analytics to provide AI agents with real-time, contextualized information. We couldn't agree more. Our award-winning Ursa Engine and Orca Agent Engine are aimed at extending our platform up the stack—bridging the gap between data streams and data lakes, and between event streams and intelligent processing.",[48,195,196],{},"Our new Ursa Engine introduces a lakehouse-native approach to streaming: it can write events directly to table formats like Iceberg on cloud storage, eliminating entire classes of ETL jobs and making fresh data instantly available for analytics queries. By integrating streaming and lakehouse technologies, we help customers collapse data silos and accelerate their AI\u002FML pipelines.",[48,198,199,200,203],{},"Beyond analytics integration, we are also enhancing StreamNative with more out-of-the-box processing and governance capabilities. In the coming months, we plan to introduce new features for lightweight stream processing and transformation, making it easier to build reactive applications directly on the platform. We're also expanding our ecosystem of connectors and integrations, so that whether your data lands in Snowflake, Databricks, or an AI model, StreamNative will seamlessly feed it. ",[44,201,202],{},"We're investing significantly in enterprise features including security, schema registry, governance, and monitoring tooling","—capabilities that are essential for mission-critical deployments and where we're committed to continued improvement.",[48,205,206],{},"This recognition from Forrester energizes us to keep innovating at full speed. We're sharing this honor with our amazing customers, community, and partners who drive us forward every day. Your feedback and real-world challenges have helped shape StreamNative into what it is today, and together, we will shape the future of streaming data. Thank you for joining us on this journey—we're just getting started, and we can't wait to deliver even more value as we continue to evolve our platform. Onward to real-time everything!",[208,209],"hr",{},[32,211,213],{"id":212},"streamnative-in-the-forrester-wave-evaluation-findings",[44,214,215,216,223],{},"StreamNative in ",[44,217,218],{},[55,219,220],{"href":57},[44,221,222],{},"The Forrester Wave™",": Evaluation Findings",[225,226,228],"h5",{"id":227},"recognized-as-a-contender-among-15-streaming-data-platform-providers","• Recognized as a Contender among 15 streaming data platform providers",[225,230,232],{"id":231},"received-the-highest-scores-possible-50-in-both-the-messaging-and-resource-optimization-criteria","* Received the highest scores possible (5.0) in both the Messaging and Resource Optimization criteria",[225,234,236],{"id":235},"cited-as-the-primary-platform-for-enterprises-wishing-to-implement-pulsar","• Cited as the primary platform for enterprises wishing to implement Pulsar",[225,238,240],{"id":239},"noted-for-excelling-at-messaging-and-resource-optimization","• Noted for excelling at messaging and resource optimization",[225,242,244],{"id":243},"customers-cited-lower-infrastructure-costs-and-excellent-support-responsiveness","• Customers cited lower infrastructure costs and excellent support responsiveness",[225,246,248],{"id":247},"recognized-for-supporting-event-driven-architectures-with-robust-scalability","• Recognized for supporting event-driven architectures with robust scalability",[208,250],{},[252,253,255,256,259,260,190],"h6",{"id":254},"forrester-disclaimer-forrester-does-not-endorse-any-company-product-brand-or-service-included-in-its-research-publications-and-does-not-advise-any-person-to-select-the-products-or-services-of-any-company-or-brand-based-on-the-ratings-included-in-such-publications-information-is-based-on-the-best-available-resources-opinions-reflect-judgment-at-the-time-and-are-subject-to-change-for-more-information-read-about-forresters-objectivity-here","**Forrester Disclaimer: **",[36,257,258],{},"Forrester does not endorse any company, product, brand, or service included in its research publications and does not advise any person to select the products or services of any company or brand based on the ratings included in such publications. Information is based on the best available resources. Opinions reflect judgment at the time and are subject to change",". *For more information, read about Forrester’s objectivity *",[55,261,265],{"href":262,"rel":263},"https:\u002F\u002Fwww.forrester.com\u002Fabout-us\u002Fobjectivity\u002F",[264],"nofollow",[36,266,267],{},"here",[208,269],{},[252,271,273],{"id":272},"apache-apache-pulsar-apache-kafka-apache-flink-and-other-names-are-trademarks-of-the-apache-software-foundation-no-endorsement-by-apache-or-other-third-parties-is-implied",[36,274,275],{},"Apache®, Apache Pulsar®, Apache Kafka®, Apache Flink® and other names are trademarks of The Apache Software Foundation. No endorsement by Apache or other third parties is implied.",{"title":18,"searchDepth":19,"depth":19,"links":277},[278,280,281,282,283,284,285],{"id":34,"depth":279,"text":38},3,{"id":42,"depth":19,"text":46},{"id":83,"depth":19,"text":86},{"id":111,"depth":19,"text":114},{"id":136,"depth":19,"text":139},{"id":165,"depth":19,"text":168},{"id":180,"depth":19,"text":183,"children":286},[287],{"id":212,"depth":279,"text":288},"StreamNative in The Forrester Wave™: Evaluation Findings","Company",null,"2025-12-16","StreamNative is recognized in The Forrester Wave™: Streaming Data Platforms, Q4 2025. Discover why Forrester highlights StreamNative's high-performance messaging, efficient resource use, and cost-effective Kafka API compatibility for real-time innovation.","\u002Fimgs\u002Fblogs\u002F693bd36cf01b217dcb67278f_Streamnative_blog_thumbnail.png",false,{},0,"\u002Fblog\u002Fstreamnative-recognized-in-the-forrester-wave-streaming-data-platforms-2025","10 mins read",{"title":26,"description":292},"blog\u002Fstreamnative-recognized-in-the-forrester-wave-streaming-data-platforms-2025",[302,303,304],"Announcements","Real-Time","Forrester","sOeeJtEO3O-IIfTPJjY1AFOMawZ_rf8FOH8A98NEKgU",{"id":307,"title":308,"authors":309,"body":311,"category":503,"createdAt":290,"date":504,"description":505,"extension":8,"featured":294,"image":506,"isDraft":294,"link":290,"meta":507,"navigation":7,"order":296,"path":508,"readingTime":509,"relatedResources":290,"seo":510,"stem":511,"tags":512,"__hash__":515},"blogs\u002Fblog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud.md","600K Topics Per Cluster: Stability Optimization of Apache Pulsar at Tencent Cloud",[310],"Xiaolong Ran",{"type":15,"value":312,"toc":497},[313,317,325,328,332,335,338,341,348,351,354,357,360,363,369,382,385,396,399,405,408,420,424,427,430,433,441,444,452,459,463,466,469,475,478,488,491],[40,314,316],{"id":315},"background","Background",[48,318,319,324],{},[55,320,323],{"href":321,"rel":322},"https:\u002F\u002Fwww.tencentcloud.com\u002F",[264],"Tencent Cloud"," is a secure, reliable, and high-performance cloud computing service provided by Tencent, one of the largest Internet companies in China and beyond. With a worldwide network of data centers, Tencent Cloud is committed to offering industry-leading solutions that integrate its cloud computing, big data, artificial intelligence, Internet of Things, security, and other advanced technologies to support the digital transformation of enterprises around the world. Tencent Cloud has 70 availability zones in 26 geographical regions, serving millions of customers from more than 100 countries and regions.",[48,326,327],{},"Currently, a Pulsar cluster at Tencent Cloud can serve around 600,000 topics in production with cost controlled at a relatively low level for different use cases. In this blog post, I will share some of our practices of optimizing Apache Pulsar for better stability and performance over the past year.",[40,329,331],{"id":330},"how-to-avoid-acknowledgment-holes","How to avoid acknowledgment holes",[48,333,334],{},"Different from other messaging systems, Pulsar supports both individual acknowledgments and cumulative acknowledgments (the latter is similar to Kafka offsets). Although individual message acknowledgments provide solutions to some online business scenarios, they also lead to acknowledgment holes.",[48,336,337],{},"Acknowledgment holes refer to the gaps between ranges, which result in fragmented acknowledgments. They are very common when you use shared subscriptions or choose to acknowledge messages individually. Pulsar uses an abstraction called individuallyDeletedMessages to track fragmented acknowledgments in the form of ranges (intervals). Essentially, this attribute is a collection of open and closed intervals. A square bracket means the message has been processed while a parenthesis indicates an acknowledgment hole.",[48,339,340],{},"In Figure 1, for example, in the first interval (5:1226..5:1280], 5 is the Ledger ID and 1226 and 1280 are the Entry IDs. As the interval is left-open and right-closed, it means 5:1280 is acknowledged and that 5:1226 is not.",[48,342,343],{},[344,345],"img",{"alt":346,"src":347},"example acknowledgment holes","\u002Fimgs\u002Fblogs\u002F63b543cad3c4c2cddd72a089_image1-221020.jpeg",[48,349,350],{},"There are many factors that can cause fragmented acknowledgments, such as the broker’s failure to process messages. In the early versions of Pulsar, there were no returns for acknowledgments, so we couldn’t ensure the acknowledgment request was correctly handled. In Apache Pulsar 2.8.0 and later versions, AckResponse was introduced for transaction messages to support returns. Another major cause is the client’s failure to call acknowledgments for some reason, which is very common in production.",[48,352,353],{},"To avoid acknowledgment holes, I listed the following two solutions that we tried at Tencent Cloud for your reference.",[48,355,356],{},"First, carefully configure the backlog size. In Pulsar, a message can be either a batch message or a single message. For a batch message, you don’t know the exact number of entries contained in it. Note that Pulsar parses batch messages on the consumer side instead of the broker side. In practice, however, it is rather difficult to precisely calculate the backlog size.",[48,358,359],{},"Second, create a broker compensatory mechanism for unacknowledged messages. As individuallyDeletedMessages contains information on unacknowledged messages, we can let the broker redeliver them to the client to fill the gaps.",[48,361,362],{},"Before I explain the details of the second solution, let’s take a look at the different stages in which messages can be in a topic. In Figure 2, a producer publishes messages on a topic, which are then received by a consumer. Messages in different states are marked in three colors.",[48,364,365],{},[344,366],{"alt":367,"src":368},"Figure 2","\u002Fimgs\u002Fblogs\u002F63b543cab14a8443e161a10e_image2-221020.png",[370,371,372,376,379],"ul",{},[373,374,375],"li",{},"Red: The latest messages sent to the topic.",[373,377,378],{},"Gray: The messages sent to the topic but not consumed by the consumer.",[373,380,381],{},"Blue: The messages already consumed and acknowledged.",[48,383,384],{},"Pulsar allows you to configure backlog policies to manage unacknowledged messages when the backlog size is exceeded.",[370,386,387,390,393],{},[373,388,389],{},"producer_exception: The broker disconnects from the client by throwing an exception. This tells the producer to stop sending new messages. It is the major policy that we are using in production at Tencent Cloud.",[373,391,392],{},"producer_request_hold: The broker holds and does not persist the producer's request payload. The producer will stop sending new messages.",[373,394,395],{},"consumer_backlog_eviction: The broker discards the oldest unacknowledged messages in the backlog to make sure the consumer can receive new messages. As messages are lost in this way, we haven’t used the policy in production.",[48,397,398],{},"So, how does Pulsar define the backlog size? In Figure 3, all messages in the stream have been consumed but not all of them have been acknowledged by the consumer.",[48,400,401],{},[344,402],{"alt":403,"src":404},"table-zk-path-topics","\u002Fimgs\u002Fblogs\u002F63b544ec6526935728445e14_table-zk-path-topics.webp",[48,406,407],{},"The Pulsar community merged some code to fix the leak issue in Pulsar 2.8+. If you are using earlier versions, you might have some dirty data in your cluster. To clean up the data, we proposed the following solution.",[409,410,411,414,417],"ol",{},[373,412,413],{},"Get a topic list through the ZooKeeper client (You can use it to read these paths and form topic names in a set format).",[373,415,416],{},"Use pulsar-admin to check whether these topics exist in the cluster. If they do not exist, the associated data must be dirty and should be deleted.",[373,418,419],{},"Keep in mind that you need to back up the data before the clean-up so that you can recover topics in case of any unexpected deletion.",[40,421,423],{"id":422},"bookie-ledger-leaks","Bookie ledger leaks",[48,425,426],{},"In production, all our retention policies are no more than 15 days. Even if we add the TTL period (for example, also 15 days), the maximum message lifecycle should be 30 days. However, we found that some ledgers which were created 2 years ago still existed and could not be deleted (We are using an internal monitoring service that checks all ledger files on a regular basis).",[48,428,429],{},"One possible reason for orphan ledgers could be the bookie CLI commands. For example, when we use some CLI commands to check the status of a cluster, it may create a ledger on the bookie. However, the retention policy is not applicable to such ledgers.",[48,431,432],{},"To delete orphan ledgers, you can try the following ways:",[409,434,435,438],{},[373,436,437],{},"Obtain the metadata of the ledger. Each ledger has its own LedgerInfo, which stores its metadata, such as the creation time and the bookies that store the ledger data. If the ledger metadata are already missing, you can delete their corresponding ledgers directly.",[373,439,440],{},"As a Pulsar topic represents a sequence of ledgers, you can check whether a ledger still exists in the ledger list of the topic. If it does not exist, you can delete it.",[48,442,443],{},"When you try to delete orphan ledgers, you need to:",[370,445,446,449],{},[373,447,448],{},"Pay special attention to the schema, which is mapped to a ledger in Pulsar. The schema ledger is stored on the bookie and the information about the schema itself is stored in ZooKeeper. If you delete the schema by accident, you need to delete the schema information on the broker first and try to recreate it from the producer side.",[373,450,451],{},"Back up your data first before you delete them.",[48,453,454,455,190],{},"For more information about how to deal with orphan ledgers, see ",[55,456,458],{"href":457},"\u002Fblog\u002Fengineering\u002F2022-09-27-a-deep-dive-into-topic-data-lifecycle-in-apache-pulsar\u002F","A Deep Dive into the Topic Data Lifecycle in Apache Pulsar",[40,460,462],{"id":461},"cache-optimization","Cache optimization",[48,464,465],{},"Pulsar uses caches at different levels. Topics have their own caches on the broker side. Write caches and read caches in BookKeeper are allocated based on JVM direct memory (25% of direct memory). For hot data, generally, these caches can be hit and there is no need to read the actual data.",[48,467,468],{},"Figure 6 shows some cache metrics we observed in one of our production cases. There was a sharp decrease of read cache size, which led to the sudden increase of read cache misses. As a result, the reads on bookies saw a peak at 16:15 with the latency increasing to nearly 5 seconds. In fact, we noticed that this sudden peak happened periodically.",[48,470,471],{},[344,472],{"alt":473,"src":474},"some graph to illustrate Cache optimization","\u002Fimgs\u002Fblogs\u002F63b544fd6526936acb446bcc_image6-221020.png",[48,476,477],{},"Let’s take a look at the following two source code snippets to analyze the reason for the above scenario.",[479,480,485],"pre",{"className":481,"code":483,"language":484},[482],"language-text","try {\n     \u002F\u002F We need to check all the segments, starting from the current\n     \u002F\u002F backward to minimize the\n     \u002F\u002F checks for recently inserted entries\n     int size = cacheSegments.size();\n     for (int i = 0; i \nIterate message\n\n","text",[486,487,483],"code",{"__ignoreMap":18},[48,489,490],{},"try {\n   int offset = currentSegmentOffset.getAndAdd(entrySize);\n   if (offset + entrySize > segmentSize) {\n       \u002F\u002F Rollover to next segment\n       currentSegmentIdx = (currentSegmentIdx + 1) % cacheSegments.size();  \n       currentSegment0ffset.set(alignedSize);\n       cacheIndexes.get(currentSegmentIdx).clear();\n        offset = 0;\n}",[479,492,495],{"className":493,"code":494,"language":484},[482],"\noffset + entrySize vs segmentSize\n\nThe first snippet uses a for loop to iterate messages for caches; in the second one, all caches will be cleared if the sum of offset and entrySize is larger than segmentSize. This explains the sudden decrease of read cache size. After that point, caches will be recreated.\n\nCurrently, we are using the LRU policy (OHC) to avoid sudden cache fluctuations. This is the result after our optimization:\n\n![Figure 7 graph read cache its](\u002Fimgs\u002Fblogs\u002F63b54548d4fc09382ee01ba5_image7-221020.png)\n\n## Summary\n\nIn this blog, we shared our experience of using and optimizing Apache Pulsar at Tencent Cloud for better performance and stability. Going forward, the Tencent Cloud team will continue to be an active player in the Pulsar community and work with other community members in the following aspects.\n\n- Retry policies within the client timeout period. We are thinking about creating an internal mechanism featuring multiple retries (send requests) to avoid message delivery failures.\n- Broker and bookie OOM optimization. Brokers may be out of memory when you have too many fragmented acknowledgments. For bookie OOM cases, they can be caused by different factors. For example, if one of the bookies in an ensemble has slow returns (Write Quorum =3, Ack Quorum = 2), the direct memory can never be released.\n- Bookie AutoRecovery optimization. AutoRecovery can be deployed separately or on the same machines where bookies are running. When you deploy them together, the AutoRecovery process can’t be restarted if you have a ZooKeeper session timeout. This is because there is no retry logic between AutoRecovery and ZooKeeper. Hence, we want to add an internal retry mechanism for AutoRecovery.\n",[486,496,494],{"__ignoreMap":18},{"title":18,"searchDepth":19,"depth":19,"links":498},[499,500,501,502],{"id":315,"depth":19,"text":316},{"id":330,"depth":19,"text":331},{"id":422,"depth":19,"text":423},{"id":461,"depth":19,"text":462},"Apache Pulsar","2022-10-20","Learn how Tencent Cloud engineers worked to optimize Apache Pulsar for better performance and stability.","\u002Fimgs\u002Fblogs\u002F63c7c146d5ec62f566f6dc78_63b543ca652693122243ce3f_apache-pulsar-at-tencent-cloud-top.jpeg",{},"\u002Fblog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud","9 min read",{"title":308,"description":505},"blog\u002F600k-topics-per-cluster-stability-optimization-apache-pulsar-tencent-cloud",[513,503,514],"Success Stories","Security","CiW_oPwNI5aoKVqR9CHs6E6pRo0YGj-Gk8sNTaq2L0M",{"id":517,"title":518,"authors":519,"body":524,"category":1280,"createdAt":290,"date":1281,"description":1282,"extension":8,"featured":7,"image":1283,"isDraft":294,"link":290,"meta":1284,"navigation":7,"order":296,"path":1285,"readingTime":1286,"relatedResources":290,"seo":1287,"stem":1288,"tags":1289,"__hash__":1292},"blogs\u002Fblog\u002Fhow-we-run-a-5-gb-s-kafka-workload-for-just-50-per-hour.md","How We Run a 5 GB\u002Fs Kafka Workload for Just $50 per Hour",[520,521,522,523],"Matteo Meril","Neng Lu","Hang Chen","Penghui Li",{"type":15,"value":525,"toc":1250},[526,529,532,535,538,541,545,548,556,561,564,572,577,581,588,591,594,602,606,609,614,618,621,624,627,630,639,643,646,657,660,664,667,670,681,684,688,692,700,703,707,715,744,748,751,756,760,763,767,770,773,778,787,792,795,798,809,813,816,827,831,834,837,842,845,874,878,880,886,889,894,899,902,906,920,924,935,939,954,963,974,977,980,984,987,990,1001,1004,1007,1010,1015,1020,1024,1028,1045,1049,1063,1068,1072,1083,1086,1102,1106,1117,1122,1127,1135,1139,1142,1146,1153,1157,1160,1169,1174,1183,1189,1198,1207,1216,1225,1234,1242],[48,527,528],{},"The rise of DeepSeek has shaken the AI infrastructure market, forcing companies to confront the escalating costs of training and deploying AI models. But the real pressure point isn’t just compute—it’s data acquisition and ingestion costs.",[48,530,531],{},"As businesses rethink their AI cost-containment strategies, real-time data streaming is emerging as a critical enabler. The growing adoption of Kafka as a standard protocol has expanded cost-efficient options, allowing companies to optimize streaming analytics while keeping expenses in check.",[48,533,534],{},"Ursa, the data streaming engine powering StreamNative’s managed Kafka service, is built for this new reality. With its leaderless architecture and native lakehouse storage integration, Ursa eliminates costly inter-zone network traffic for data replication and client-to-broker communication while ensuring high availability at minimal operational cost.",[48,536,537],{},"In this blog post, we benchmarked the infrastructure cost and total cost of ownership (TCO) for running a 5GB\u002Fs Kafka workload across different Kafka vendors, including Redpanda, Confluent WarpStream, and AWS MSK. Our benchmark results show that Ursa can sustain 5GB\u002Fs Kafka workloads at just 5% of the cost of traditional streaming engines like Redpanda—making it the ideal solution for high-performance, cost-efficient ingestion and data streaming for data lakehouses and AI workloads.",[48,539,540],{},"Note: We also evaluated vanilla Kafka in our benchmark; however, for simplicity, we have focused our cost comparison on vendor solutions rather than self-managed deployments. That said, it is important to highlight that both Redpanda and vanilla Kafka use a leader-based data replication approach. In a data-intensive, network-bound workload like 5GB\u002Fs streaming, with the same machine type and replication factor, Redpanda and vanilla Kafka produced nearly identical cost profiles.",[40,542,544],{"id":543},"key-benchmark-findings","Key Benchmark Findings",[48,546,547],{},"Ursa delivered 5 GB\u002Fs of sustained throughput at an infrastructure cost of just $54 per hour. For comparison:",[370,549,550,553],{},[373,551,552],{},"MSK: $303 per hour → 5.6x more expensive compared to Ursa",[373,554,555],{},"Redpanda: $988 per hour → 18x more expensive compared to Ursa",[48,557,558],{},[344,559],{"alt":18,"src":560},"\u002Fimgs\u002Fblogs\u002F679c71b67d9046f26edc7977_AD_4nXfvTqyBNUBu2lObdkKAx-5UNkpNP8UYULLZyOcixE6z99VMZUUEsUqWjzexI7vjyNGRNSAUoM9smYvdTP55ctAhIbrs5lmQgcSVMWdaoigbWouCl95DVSQsxooY-qqfGcYqS4g4zA.png",[48,562,563],{},"Beyond infrastructure costs, when factoring in both storage pricing, vendor pricing and operational expenses, Ursa’s total cost of ownership (TCO) for a 5GB\u002Fs workload with a 7-day retention period is:",[370,565,566,569],{},[373,567,568],{},"50% cheaper than Confluent WarpStream",[373,570,571],{},"85% cheaper than MSK and Redpanda",[48,573,574],{},[344,575],{"alt":18,"src":576},"\u002Fimgs\u002Fblogs\u002F679c602d77e9c706de5343b8_AD_4nXeDv8rrv_C1CTCCiqYo1zpvlGYbdBk1r0VEqovAPu22iFMQZgh54Hfw9PBMLzM7jDFxKwAFDxbdG0np4XVk_tGsWhEKMloLRcmmea7lvueCx-0cFsyaE3Mya4Mxc1Dox95A6JEc.png",[40,578,580],{"id":579},"ursa-highly-cost-efficient-data-streaming-at-scale","Ursa: Highly Cost-Efficient Data Streaming at Scale",[48,582,583,587],{},[55,584,586],{"href":585},"\u002Fblog\u002Fursa-reimagine-apache-kafka-for-the-cost-conscious-data-streaming","Ursa"," is a next-generation data streaming engine designed to deliver high performance at a fraction of the cost of traditional disk-based solutions. It is fully compatible with Apache Kafka and Apache Pulsar APIs, while leveraging a leaderless, lakehouse-native architecture to maximize scalability, efficiency, and cost savings.",[48,589,590],{},"Ursa’s key innovation is separating storage from compute and decoupling metadata\u002Findex operations from data operations by utilizing cloud object storage (e.g., AWS S3) instead of costly inter-zone disk-based replication. It also employs open lakehouse formats (Iceberg and Delta Lake), enabling columnar compression to significantly reduce storage costs while maintaining durability and availability.",[48,592,593],{},"In contrast, traditional streaming systems—like Kafka and Redpanda—depend on leader-based architectures, which drive up inter-zone traffic costs due to replication and client communication. Ursa mitigates these costs by:",[370,595,596,599],{},[373,597,598],{},"Eliminating inter-zone traffic costs via a leaderless architecture.",[373,600,601],{},"Replacing costly inter-zone replication with direct writes to cloud storage using open lakehouse formats.",[40,603,605],{"id":604},"how-ursa-eliminates-inter-zone-traffic","How Ursa Eliminates Inter-Zone Traffic",[48,607,608],{},"Ursa minimizes inter-zone traffic by leveraging a leaderless architecture, which eliminates inter-zone communication between clients and brokers, and lakehouse-native storage, which removes the need for inter-zone data replication. This approach ensures high availability and scalability while avoiding unnecessary cross-zone data movement.",[48,610,611],{},[344,612],{"alt":18,"src":613},"\u002Fimgs\u002Fblogs\u002F679c602e21b3571bb7117dca_AD_4nXd7Oahc77NjRLNvA9clLt0tsyU6MrIqVibFYv5pW5giTIcCHPr3EA_yTGzfVEUIVO3VXK56qWK8zmBCp5lY0E_4nmlWIPFrHjtHylA5NhwELjn-UB0fLG2h_kbrxrc7Cs_edvveNA.png",[32,615,617],{"id":616},"leaderless-architecture","Leaderless architecture",[48,619,620],{},"Traditional streaming engines such as Kafka, Pulsar, or RedPanda rely on a leader-based model, where each partition is assigned to a single leader broker that handles all writes and reads.",[48,622,623],{},"Pros of Leader-Based Architectures:\n✔ Maintains message ordering via local sequence IDs\n✔ Delivers low latency and high performance through message caching",[48,625,626],{},"Cons of Leader-Based Architectures:\n✖ Throughput bottlenecked by a single broker per partition\n✖ Inter-zone traffic required for high availability in multi-AZ deployments",[48,628,629],{},"While Kafka and Pulsar offer partial solutions (e.g., reading from followers, shadow topics) to reduce read-related inter-zone traffic, producers still send data to a single leader.",[48,631,632,633,638],{},"Ursa removes the concept of topic ownership, allowing any broker in the cluster to handle reads or writes for any partition. The primary challenge—ensuring message ordering—is solved with ",[55,634,637],{"href":635,"rel":636},"https:\u002F\u002Fgithub.com\u002Fstreamnative\u002Foxia",[264],"Oxia",", a scalable metadata and index service created by StreamNative in 2022.",[32,640,642],{"id":641},"oxia-the-metadata-layer-enabling-leaderless-architecture","Oxia: The Metadata Layer Enabling Leaderless Architecture",[48,644,645],{},"Ensuring message ordering in a leaderless architecture is complex, but Ursa solves this with Oxia:",[370,647,648,651,654],{},[373,649,650],{},"Handles millions of metadata\u002Findex operations per second",[373,652,653],{},"Generates sequential IDs to maintain strict message ordering",[373,655,656],{},"Optimized for Kubernetes with horizontal scalability",[48,658,659],{},"Producers and consumers can connect to any broker within their local AZ, eliminating inter-zone traffic costs while maintaining performance through localized caching.",[32,661,663],{"id":662},"zero-interzone-data-replication","Zero interzone data replication",[48,665,666],{},"In most distributed systems, data replication from a leader (primary) to followers (replicas) is crucial for fault tolerance and availability. However, replication across zones can inflate infrastructure expenses substantially.",[48,668,669],{},"Ursa avoids these costs by writing data directly to cloud storage (e.g., AWS S3, Google GCS):",[370,671,672,675,678],{},[373,673,674],{},"Built-In Resilience: Cloud storage inherently offers high availability and fault tolerance without inter-zone traffic fees.",[373,676,677],{},"Tradeoff: Slightly higher latency (sub-second, with p99 at 500 milliseconds) compared to local disk\u002FEBS (single-digit to sub-100 milliseconds), in exchange for significantly lower costs (up to 10x lower).",[373,679,680],{},"Flexible Modes: Ursa is an addition to the classic BookKeeper-based engine, providing users with the flexibility to optimize for either cost or low latency based on their workload requirements.",[48,682,683],{},"By foregoing conventional replication, Ursa slashes inter-zone traffic costs and associated complexities—making it a compelling option for organizations seeking to balance high-performance data streaming with strict budget constraints.",[40,685,687],{"id":686},"how-we-ran-a-5-gbs-test-with-ursa","How We Ran a 5 GB\u002Fs Test with Ursa",[32,689,691],{"id":690},"ursa-cluster-deployment","Ursa Cluster Deployment",[370,693,694,697],{},[373,695,696],{},"9 brokers across 3 availability zones, each on m6i.8xlarge (Fixed 12.5 Gbps bandwidth, 32 vCPU cores, 128 GB memory).",[373,698,699],{},"Oxia cluster (metadata store) with 3 nodes of m6i.8xlarge, distributed across three availability zones (AZs).",[48,701,702],{},"During peak throughput (5 GB\u002Fs), each broker’s network usage was about 10 Gbps.",[32,704,706],{"id":705},"openmessaging-benchmark-workers-configuration","OpenMessaging Benchmark Workers & Configuration",[48,708,709,710,714],{},"The OpenMessaging Benchmark(OMB) Framework is a suite of tools that make it easy to benchmark distributed messaging systems in the cloud. Please check ",[55,711,712],{"href":712,"rel":713},"https:\u002F\u002Fopenmessaging.cloud\u002Fdocs\u002Fbenchmarks\u002F",[264]," for details.",[370,716,717,732,741],{},[373,718,719,720,725,726,731],{},"12 OMB workers: 6 for ",[55,721,724],{"href":722,"rel":723},"https:\u002F\u002Fgist.github.com\u002Fcodelipenghui\u002Fd1094122270775e4f1580947f80c5055",[264],"producers",", 6 for ",[55,727,730],{"href":728,"rel":729},"https:\u002F\u002Fgist.github.com\u002Fcodelipenghui\u002F06bada89381fb77a7862e1b4c1d8963d",[264],"consumers"," across 3 availability zones, on m6i.8xlarge instances. Each worker is configured with 12 CPU cores and 48 GB memory.",[373,733,734,735,740],{},"Sample YAML ",[55,736,739],{"href":737,"rel":738},"https:\u002F\u002Fgist.github.com\u002Fcodelipenghui\u002F204c1f26c4d44a218ae235bf2de99904",[264],"scripts"," provided for Kafka-compatible configuration and rate limits.",[373,742,743],{},"Achieved consistent 5 GB\u002Fs publish\u002Fsubscribe throughput.",[40,745,747],{"id":746},"ursa-benchmark-tests-results","Ursa Benchmark Tests & Results",[48,749,750],{},"The following diagram demonstrates that Ursa can consistently handle 5 GB\u002Fs of traffic, fully saturating the network across all broker nodes.",[48,752,753],{},[344,754],{"alt":18,"src":755},"\u002Fimgs\u002Fblogs\u002F679c602d7b261bac1113f7d6_AD_4nXdDPsRc3koXICiFF0bqSmGWbJt_RlUy4FE3ruuWOfbCfpcqZ1dejjqGbkaCJv2hQFL1nirRouBVRW2l5uMWBvY9naMqGB_wHcLI14dBM0f85TXhmdm3UxEv1yGX9Y4hf5FttSkZew.png",[40,757,759],{"id":758},"comparing-infrastructure-cost","Comparing Infrastructure Cost",[48,761,762],{},"This benchmark first evaluates infrastructure costs of running a 5 GB\u002Fs streaming workload (1:1 producer-to-consumer ratio) across different data streaming engines, including Ursa, Redpanda, and AWS MSK, with a focus on multi-AZ deployments to ensure a fair comparison.",[32,764,766],{"id":765},"test-setup-key-assumptions","Test Setup & Key Assumptions",[48,768,769],{},"All tests use multi-AZ configurations, with clusters and clients distributed across three AWS availability zones (AZs). Cluster size scales proportionally to the number of AZs, and rack-awareness is enabled for all engines to evenly distribute topic partitions and leaders.",[48,771,772],{},"To ensure a fair comparison, we selected the same machine type capable of fully utilizing both network and storage bandwidth for Ursa and Redpanda in this 5GB\u002Fs test:",[370,774,775],{},[373,776,777],{},"9 × m6i.8xlarge instances",[48,779,780,781,786],{},"However, MSK's storage bandwidth limits vary depending on the selected instance type, with the highest allowed limit capped at 1000 MiB\u002Fs per broker, according to",[55,782,785],{"href":783,"rel":784},"https:\u002F\u002Fdocs.aws.amazon.com\u002Fmsk\u002Flatest\u002Fdeveloperguide\u002Fmsk-provision-throughput-management.html#throughput-bottlenecks",[264]," AWS documentation",". Given this constraint, achieving 5 GB\u002Fs throughput with a replication factor of 3 required the following setup:",[370,788,789],{},[373,790,791],{},"15 × kafka.m7g.8xlarge (32 vCPUs, 128 GB memory, 15 Gbps network, 4000 GiB EBS).",[48,793,794],{},"This configuration was necessary to work around MSK's storage bandwidth limitations, ensuring a comparable cost basis to other evaluated streaming engines.",[48,796,797],{},"Additional key assumptions include:",[370,799,800,803,806],{},[373,801,802],{},"Inter-AZ producer traffic: For leader-based engines, two-thirds of producer-to-broker traffic crosses AZs due to leader distribution.",[373,804,805],{},"Consumer optimizations: Follower fetch is enabled across all tests, eliminating inter-AZ consumer traffic.",[373,807,808],{},"Storage cost exclusions: This benchmark only evaluates streaming costs, assuming no long-term data retention.",[32,810,812],{"id":811},"inter-broker-replication-costs","Inter-Broker Replication Costs",[48,814,815],{},"Inter-broker (cross-AZ) replication is a major cost driver for data streaming engines:",[370,817,818,821,824],{},[373,819,820],{},"RedPanda: Inter-broker replication is not free, leading to substantial costs when data must be copied across multiple availability zones.",[373,822,823],{},"AWS MSK: Inter-broker replication is free, but MSK instance pricing is significantly higher (e.g., $3.264 per hour for kafka.m7g.8xlarge vs $1.306 per hour for an on-demand m7g.8xlarge). The storage price of MSK is $0.10 per GB-month which is significantly higher than st1, which costs $0.045 per GB-month. Even though replication is free, client-to-broker traffic still incurs inter-AZ charges.",[373,825,826],{},"Ursa: No inter-broker replication costs due to its leaderless architecture, eliminating inter-zone replication costs entirely.",[32,828,830],{"id":829},"zone-affinity-reducing-inter-az-costs","Zone Affinity: Reducing Inter-AZ Costs",[48,832,833],{},"We evaluated zone affinity mechanisms to further reduce inter-AZ data transfer costs.",[48,835,836],{},"Consumers:",[370,838,839],{},[373,840,841],{},"Follower fetch is enabled across all tests, ensuring consumers fetch data from replicas in their local AZ—eliminating inter-zone consumer traffic except for metadata lookups",[48,843,844],{},"Producers:",[370,846,847,856,865],{},[373,848,849,850,855],{},"Kafka protocol lacks an easy way to enforce producer AZ affinity (though ",[55,851,854],{"href":852,"rel":853},"https:\u002F\u002Fcwiki.apache.org\u002Fconfluence\u002Fdisplay\u002FKAFKA\u002FKIP-1123:+Rack-aware+partitioning+for+Kafka+Producer",[264],"KIP-1123"," aims to address this). And it only works with the default partitioner (i.e., when no record partition or record key is specified).",[373,857,858,859,864],{},"Redpanda recently introduced ",[55,860,863],{"href":861,"rel":862},"https:\u002F\u002Fdocs.redpanda.com\u002Fredpanda-cloud\u002Fdevelop\u002Fproduce-data\u002Fleader-pinning\u002F",[264],"leader pinning",", but this only benefits setups where producers are confined to a single AZ—not applicable to our multi-AZ benchmark.",[373,866,867,868,873],{},"Ursa is the only system in this test with ",[55,869,872],{"href":870,"rel":871},"https:\u002F\u002Fdocs.streamnative.io\u002Fdocs\u002Fconfig-kafka-client#eliminate-cross-az-networking-traffic",[264],"built-in zone affinity for both producers and consumers",". It achieves this by embedding producer AZ information in client.id, allowing metadata lookups to route clients to local-AZ brokers, eliminating inter-AZ producer traffic.",[32,875,877],{"id":876},"cost-comparison-results","Cost Comparison Results",[48,879,547],{},[370,881,882,884],{},[373,883,552],{},[373,885,555],{},[48,887,888],{},"Ursa’s leaderless architecture, zone affinity, and native cloud storage integration deliver unparalleled cost efficiency, making it the most cost-effective choice for high-throughput data streaming workloads.",[48,890,891],{},[344,892],{"alt":18,"src":893},"\u002Fimgs\u002Fblogs\u002F679c72208198ca36a352f228_AD_4nXeeZuM8T-xBlD4Vf3j67K618n08qh8wIDLLtiLJG0ssA1Wj1V26u7wIDTX9sqLrtw8mB2c299dwzarGen62CG0Vh7nWstn5qbPGFcBaKJYEepTsLr5fHWv1U8uqbg8Y0UOK6fJ7.png",[48,895,896],{},[344,897],{"alt":18,"src":898},"\u002Fimgs\u002Fblogs\u002F679c625978031f40229de484_AD_4nXdLkLLJ30KKr-_A_rN1j8akVwBYacAWIPzWHoOReJF421890kfByZoQQxkLczihVSmiw5Q9J51-V9I2SEKITbwsYnANDDTlAVL5nQ_jfaHNTe9VEWhSoa7DZooCnilDYL6l6msmJg.png",[48,900,901],{},"The detailed infrastructure cost calculations for each data streaming engine are listed below:",[32,903,905],{"id":904},"streamnative-ursa","StreamNative - Ursa",[370,907,908,911,914,917],{},[373,909,910],{},"Server EC2 costs: 9 * $1.536\u002Fhr = $14",[373,912,913],{},"Client EC2 costs: 9 * $1.536\u002Fhr =$14",[373,915,916],{},"S3 write requests costs: 1350 r\u002Fs * $0.005\u002F1000r * 3600s = $24",[373,918,919],{},"S3 read requests costs: 1350 r\u002Fs * $0.0004\u002F1000r * 3600s = $2",[32,921,923],{"id":922},"aws-msk","AWS MSK",[370,925,926,929,932],{},[373,927,928],{},"Server EC2 costs: 15 * $3.264\u002Fhr = $49",[373,930,931],{},"Client side EC2 costs: 9 * $1.536\u002Fhr =$14",[373,933,934],{},"Interzone traffic - producer to broker: 5GB\u002Fs * ⅔ * $0.02\u002FG(in+out) * 3600 = $240",[32,936,938],{"id":937},"redpanda","RedPanda",[370,940,941,943,945,948,951],{},[373,942,910],{},[373,944,913],{},[373,946,947],{},"Interzone traffic - producer to broker: 5GB\u002Fs * ⅔ * $0.02\u002FGB(in+out) * 3600 = $240",[373,949,950],{},"Interzone traffic - replication: 10GB\u002Fs * $0.02\u002FGB(in+out) * 3600 = $720",[373,952,953],{},"Interzone traffic - broker to consumer: $0 (fetch from local zone)",[48,955,956,957,962],{},"Please note that we were unable to test ",[55,958,961],{"href":959,"rel":960},"https:\u002F\u002Fwww.redpanda.com\u002Fblog\u002Fcloud-topics-streaming-data-object-storage",[264],"Redpanda with Cloud Topics",", as it remains an announced but unreleased feature and is not yet available for evaluation. Based on the limited information available, while Cloud Topics may help optimize inter-zone data replication costs, producers still need to traverse inter-availability zones to connect to the topic partition owners and incur inter-zone traffic costs of up to $240 per hour.",[370,964,965,971],{},[373,966,967,970],{},[55,968,854],{"href":852,"rel":969},[264]," (when implemented) will help mitigate producer-to-broker inter-zone traffic, but it is not yet available. And it only works with the default partitioner (no record partition or key is specified).",[373,972,973],{},"Redpanda’s leader pinning helps only when all producers for the pinned topic are confined to a single AZ. In multi-AZ environments (like our benchmark), inter-zone producer traffic remains unavoidable.",[48,975,976],{},"Additionally, Redpanda’s Cloud Topics architecture is not documented publicly. Their blog mentions \"leader placement rules to optimize produce latency and ingress cost,\" but it is unclear whether this represents a shift away from a leader-based architecture or if it uses techniques similar to Ursa’s zone-aware approach.",[48,978,979],{},"We may revisit this comparison as more details become available.",[40,981,983],{"id":982},"comparing-total-cost-of-ownership","Comparing Total Cost of Ownership",[48,985,986],{},"As highlighted earlier, with a BYOC Ursa setup, you can achieve 5 GB\u002Fs throughput at just 5% of the infrastructure cost of a traditional leader-based data streaming engine, such as Kafka or RedPanda, while managing the infrastructure yourself. This significant cost reduction is enabled by Ursa’s leaderless architecture and lakehouse-native storage design, which eliminate overhead costs such as inter-zone traffic and leader-based data replication. By leveraging a lakehouse-native, leaderless architecture, Ursa reduces resource requirements, enabling you to handle high data throughput efficiently and at a fraction of the cost of RedPanda.",[48,988,989],{},"Now, let’s examine the total cost comparison, evaluating Ursa alongside other vendors, including those that have adopted a leaderless architecture (e.g., Confluent WarpStream). This comparison is based on a 5GB\u002Fs workload with a 7-day retention period, factoring in both storage cost and vendor costs Here are the key findings:",[370,991,992,995,998],{},[373,993,994],{},"Ursa ($164,353\u002Fmonth) is: 50% cheaper than Confluent WarpStream ($337,068\u002Fmonth)",[373,996,997],{},"85% cheaper than AWS MSK ($1,115,251\u002Fmonth)",[373,999,1000],{},"86% cheaper than Redpanda ($1,202,853\u002Fmonth)",[48,1002,1003],{},"In addition to Ursa’s architectural advantages—eliminating most inter-AZ traffic and leveraging lakehouse storage for cost-effective data retention—it also adopts a more fair and cost-efficient pricing model: Elastic Throughput-based pricing. This approach aligns costs with actual usage, avoiding unnecessary overhead.",[48,1005,1006],{},"Unlike WarpStream, which charges for both storage and throughput, Ursa ensures that customers only pay for the throughput they actively use. Ursa’s pricing is based on compressed data sent by clients, meaning the more data compressed on the client side, the lower the cost. In contrast, WarpStream prices are based on uncompressed data, unfairly inflating expenses and failing to incentivize customers to optimize their client applications.",[48,1008,1009],{},"This distinction is crucial, as compressed data reduces both storage and network costs, making Ursa’s pricing model not only more cost-effective but also more transparent and predictable.",[48,1011,1012],{},[344,1013],{"alt":18,"src":1014},"\u002Fimgs\u002Fblogs\u002F679c602d194800c9206d9d58_AD_4nXcFlf755xgyz7htxhMhBV5fGrsxy642mQNodt61DTok_z1dwkw5A6lkO5hatXVneCaB0anbZPAyvLI3MlIMuQEYLEACHHvQMOr5UfaB37dfzkdqewDEvcT-20VGd_zzvJsuA00zGA.png",[48,1016,1017],{},[344,1018],{"alt":18,"src":1019},"\u002Fimgs\u002Fblogs\u002F679c62594e9c2e629fae73aa_AD_4nXeU6cOgItnjLsEZCOf13TEvMY_SHWWIxYP2OYUj-B1GUPyWO78OG08K_v03hwYSVcg06f9dqDiGmdwy76vynjmiDGL5bluZ5_XF4nSU_r59oOZdfViXndXt6s11vVOY7qwfZN8v.png",[32,1021,1023],{"id":1022},"cost-breakdown","Cost Breakdown",[1025,1026,1027],"h4",{"id":904},"StreamNative – Ursa",[370,1029,1030,1033,1036,1039,1042],{},[373,1031,1032],{},"EC2 (Server): 9 × $1.536\u002Fhr × 24 hr × 30 days = $9,953.28",[373,1034,1035],{},"S3 Write Requests: 1,350 r\u002Fs × $0.005\u002F1,000 r × 3,600 s × 24 hr × 30 days = $17,496",[373,1037,1038],{},"S3 Read Requests: 1,350 r\u002Fs × $0.0004\u002F1,000 r × 3,600 s × 24 hr × 30 days = $1,400",[373,1040,1041],{},"S3 Storage Costs: 5 GB\u002Fs × $0.021\u002FGB × 3,600 s × 24 hr × 7 days = $63,504",[373,1043,1044],{},"Vendor Cost: 200 ETU × $0.50\u002Fhr × 24 hr × 30 days = $72,000",[1025,1046,1048],{"id":1047},"warpstream","WarpStream",[370,1050,1051,1054],{},[373,1052,1053],{},"Based on WarpStream’s pricing calculator (as of January 29, 2025), we assume a 4:1 client data compression ratio, meaning 20 GB\u002Fs of uncompressed data translates to 5 GB\u002Fs of compressed data.",[373,1055,1056,1057,1062],{},"It's important to note that WarpStream’s pricing structure has fluctuated frequently throughout January. We observed the cost reported by their calculator changing from $409,644 per month to $337,068 per month. This variability has been previously highlighted in the blog post “",[55,1058,1061],{"href":1059,"rel":1060},"https:\u002F\u002Fbigdata.2minutestreaming.com\u002Fp\u002Fthe-brutal-truth-about-apache-kafka-cost-calculators",[264],"The Brutal Truth About Kafka Cost Calculators","”. To ensure transparency, we have documented the pricing as of January 29, 2025.",[48,1064,1065],{},[344,1066],{"alt":18,"src":1067},"\u002Fimgs\u002Fblogs\u002F679c602e42713e0028e9af5e_AD_4nXcu5_VWTLu9jRYs6zX1MBAOtLQEo5gyfNSWPcbpnQHXTa8qNCFAXezRR2E8daygzYTTwd4dhJjaLaLM8C6y_3OGbu2NS7pdvEv3a8-ptNKOg7AeKnYqPQCAYvQ5EuxzuI3JYIvY.png",[1025,1069,1071],{"id":1070},"msk","MSK",[370,1073,1074,1077,1080],{},[373,1075,1076],{},"EC2 (Server): 15 * $3.264\u002Fhr × 24 hr × 30 days = $35,251",[373,1078,1079],{},"Interzone Traffic (Client-Server): 5 GB\u002Fs × ⅔ × $0.02\u002FGB (in+out) × 3,600 s × 24 hr × 30 days = $172,800",[373,1081,1082],{},"Storage: 5 GB\u002Fs × $0.1\u002FGB-month × 3,600 s × 24 hr × 7 days * 3 replicas = $907,200",[1025,1084,938],{"id":1085},"redpanda-1",[370,1087,1088,1091,1093,1096,1099],{},[373,1089,1090],{},"EC2 (Server): 9 × $1.536\u002Fhr × 24 hr × 30 days = $9953",[373,1092,1079],{},[373,1094,1095],{},"Interzone Traffic (Replication): 5 GB\u002Fs × 2 × $0.02\u002FGB (in+out) × 3,600 s × 24 hr × 30 days = $518,400",[373,1097,1098],{},"Storage: 5 GB\u002Fs × $0.045\u002FGB-month(st1) × 3,600 s × 24 hr × 7 days * 3 replicas = $408,240",[373,1100,1101],{},"Vendor Cost: $93,333 per month (based on limited information. See additional notes below).",[1025,1103,1105],{"id":1104},"additional-notes","Additional Notes",[370,1107,1108],{},[373,1109,1110,1111,1116],{},"Redpanda does not publicly disclose its BYOC pricing, making it difficult to accurately assess its total costs. We refer to information from the whitepaper “",[55,1112,1115],{"href":1113,"rel":1114},"https:\u002F\u002Fwww.redpanda.com\u002Fresources\u002Fredpanda-vs-confluent-performance-tco-benchmark-report#form",[264],"Redpanda vs. Confluent: A Performance and TCO Benchmark Report by McKnight Consulting Group.","” for estimation purposes. Based on the Tier-8 pricing model in the whitepaper,  the estimated cost to support a 5GB\u002Fs workload would be $1.12 million per year ($93,333 per month). However, since this calculation is based on an estimation, we will revisit and refine the cost assessment once Redpanda publishes its BYOC pricing.",[48,1118,1119],{},[344,1120],{"alt":18,"src":1121},"\u002Fimgs\u002Fblogs\u002F679c602dc8a9859eed89a0ef_AD_4nXdbcO8vsNNPy4GtkNLlmNKf22fjxRvzLzH7CtOna1L08sTbvnZx3HhufeFqc1w4K2gEF7lxO2IR5supotxebAiGnA07Qa8Yr3Rd1pVK2LYKK4WurlJGwgdwwucZIFoF-N_2oBjY.png",[48,1123,1124],{},[344,1125],{"alt":18,"src":1126},"\u002Fimgs\u002Fblogs\u002F679c602d6bc1c2287e012540_AD_4nXfcHZnLfjbjIr3ZAgoQXT9dwP3aQCOQPmGZZJUtpNZSwE6qY6M3yehIaBxCwxEIeu5PVdUPY0zhyjnow26YfgjdYgSG4GnV9ibxu0YWTIpwng6z_F6FUGJMpERMKtpsFESzXSN_Sw.png",[370,1128,1129,1132],{},[373,1130,1131],{},"When estimating the storage costs for Kafka and Redpanda, we assume the use of HDD storage at $0.045\u002FGB, based on the premise that both systems can fully utilize disk bandwidth without incurring the higher costs associated with GP2 or GP3 volumes. However, in practice, many users opt for GP2 or GP3, significantly increasing the total storage cost for Kafka and Redpanda.",[373,1133,1134],{},"Unlike disk-based solutions, S3 storage does not require capacity preallocation—Ursa only incurs costs for the actual data stored. This contrasts with Kafka and Redpanda, where preallocating storage can drive up expenses. As a result, the real-world storage costs for Kafka and Redpanda are often 50% higher than the estimates above.",[40,1136,1138],{"id":1137},"conclusion","Conclusion",[48,1140,1141],{},"Ursa represents a transformative shift in streaming data infrastructure, offering cost efficiency, scalability, and flexibility without compromising durability or reliability. By leveraging a leaderless architecture and eliminating inter-zone data replication, Ursa reduces total cost of ownership by over 90% compared to traditional leader-based streaming engines like Kafka and Redpanda. Its direct integration with cloud storage and scalable metadata & index management via Oxia ensure high availability and simplified infrastructure management.",[32,1143,1145],{"id":1144},"balancing-latency-and-cost","Balancing Latency and Cost",[48,1147,1148,1152],{},[55,1149,1151],{"href":1150},"\u002Fblog\u002Fcap-theorem-for-data-streaming","Ursa trades off slightly higher latency for ultra low cost",", making it an ideal choice for the majority of streaming workloads, especially those that prioritize throughput and cost savings over ultra-low latency. Meanwhile, StreamNative’s BookKeeper-based engine remains the preferred solution for real-time, latency-sensitive applications. By combining these two approaches, StreamNative empowers customers with the flexibility to choose the right engine for their specific needs—whether it's maximizing cost savings or achieving ultra low-latency real-time performance.",[32,1154,1156],{"id":1155},"the-future-of-streaming-infrastructure","The Future of Streaming Infrastructure",[48,1158,1159],{},"In an era where data fuels AI, analytics, and real-time decision-making, managing infrastructure costs is critical to sustaining innovation. Ursa is not just a cost-cutting alternative—it is a forward-thinking, lakehouse-native platform that redefines how modern data streaming infrastructure should be built and operated.",[48,1161,1162,1163,1168],{},"Whether your priority is reducing costs, improving flexibility, or ingesting massive data into lakehouses, Ursa delivers a future-proof solution for the evolving demands of real-time data streaming. ",[55,1164,1167],{"href":1165,"rel":1166},"https:\u002F\u002Fconsole.streamnative.cloud\u002F",[264],"Get started"," with StreamNative Ursa today!",[1170,1171,1173],"h1",{"id":1172},"references","References",[48,1175,1176,1179,1180],{},[1177,1178,637],"span",{}," ",[55,1181,1182],{"href":1182},"\u002Fblog\u002Fintroducing-oxia-scalable-metadata-and-coordination",[48,1184,1185,1179,1187],{},[1177,1186,586],{},[55,1188,585],{"href":585},[48,1190,1191,1179,1194],{},[1177,1192,1193],{},"StreamNative pricing",[55,1195,1196],{"href":1196,"rel":1197},"https:\u002F\u002Fdocs.streamnative.io\u002Fdocs\u002Fbilling-overview",[264],[48,1199,1200,1179,1203],{},[1177,1201,1202],{},"WarpStream pricing",[55,1204,1205],{"href":1205,"rel":1206},"https:\u002F\u002Fwww.warpstream.com\u002Fpricing#pricingfaqs",[264],[48,1208,1209,1179,1212],{},[1177,1210,1211],{},"AWS S3 pricing",[55,1213,1214],{"href":1214,"rel":1215},"https:\u002F\u002Faws.amazon.com\u002Fs3\u002Fpricing\u002F",[264],[48,1217,1218,1179,1221],{},[1177,1219,1220],{},"AWS EBS pricing",[55,1222,1223],{"href":1223,"rel":1224},"https:\u002F\u002Faws.amazon.com\u002Febs\u002Fpricing\u002F",[264],[48,1226,1227,1179,1230],{},[1177,1228,1229],{},"AWS MSK pricing",[55,1231,1232],{"href":1232,"rel":1233},"https:\u002F\u002Faws.amazon.com\u002Fmsk\u002Fpricing\u002F",[264],[48,1235,1236,1179,1239],{},[1177,1237,1238],{},"The Brutal Truth about Kafka Cost Calculators",[55,1240,1059],{"href":1059,"rel":1241},[264],[48,1243,1244,1179,1247],{},[1177,1245,1246],{},"Redpanda vs. Confluent: A Performance and TCO Benchmark Report by McKnight Consulting Group",[55,1248,1113],{"href":1113,"rel":1249},[264],{"title":18,"searchDepth":19,"depth":19,"links":1251},[1252,1253,1254,1259,1263,1264,1273,1276],{"id":543,"depth":19,"text":544},{"id":579,"depth":19,"text":580},{"id":604,"depth":19,"text":605,"children":1255},[1256,1257,1258],{"id":616,"depth":279,"text":617},{"id":641,"depth":279,"text":642},{"id":662,"depth":279,"text":663},{"id":686,"depth":19,"text":687,"children":1260},[1261,1262],{"id":690,"depth":279,"text":691},{"id":705,"depth":279,"text":706},{"id":746,"depth":19,"text":747},{"id":758,"depth":19,"text":759,"children":1265},[1266,1267,1268,1269,1270,1271,1272],{"id":765,"depth":279,"text":766},{"id":811,"depth":279,"text":812},{"id":829,"depth":279,"text":830},{"id":876,"depth":279,"text":877},{"id":904,"depth":279,"text":905},{"id":922,"depth":279,"text":923},{"id":937,"depth":279,"text":938},{"id":982,"depth":19,"text":983,"children":1274},[1275],{"id":1022,"depth":279,"text":1023},{"id":1137,"depth":19,"text":1138,"children":1277},[1278,1279],{"id":1144,"depth":279,"text":1145},{"id":1155,"depth":279,"text":1156},"StreamNative Cloud","2025-01-31","Discover how Ursa achieves 5GB\u002Fs Kafka workloads at just 5% of the cost of traditional streaming engines like Redpanda and AWS MSK. See our benchmark results comparing infrastructure costs, total cost of ownership (TCO), and performance across leading Kafka vendors.","\u002Fimgs\u002Fblogs\u002F679c6593d25099b1cdcec4ca_image-31.png",{},"\u002Fblog\u002Fhow-we-run-a-5-gb-s-kafka-workload-for-just-50-per-hour","30 min",{"title":518,"description":1282},"blog\u002Fhow-we-run-a-5-gb-s-kafka-workload-for-just-50-per-hour",[1290,1291,303],"TCO","Apache Kafka","A0o_2xdJiLI6rf6xj4RKsxJNo_A6QN2fYzCp6gaLrFw",[1294],{"id":1295,"title":310,"bioSummary":1296,"email":290,"extension":8,"image":1297,"linkedinUrl":290,"meta":1298,"position":1305,"stem":1306,"twitterUrl":1307,"__hash__":1308},"authors\u002Fauthors\u002Fxiaolong-ran.md","Xiaolong Ran is a Senior Software Engineer at Tencent Cloud, an Apache Pulsar Committer, an RoP maintainer, and an Apache Pulsar Go Client and Go Functions Developer and Maintainer.","\u002Fimgs\u002Fauthors\u002Fxiaolong-ran.webp",{"body":1299},{"type":15,"value":1300,"toc":1303},[1301],[48,1302,1296],{},{"title":18,"searchDepth":19,"depth":19,"links":1304},[],"Senior Software Engineer, Tencent Cloud","authors\u002Fxiaolong-ran","https:\u002F\u002Ftwitter.com\u002Fwolf4j1","yrPYSxkL5D9GvnZtggrbRRcnduV1hifa9K_UYRSCEGU",[1310,1317,1323],{"path":1311,"title":1312,"date":1313,"image":-1,"link":-1,"collection":1314,"resourceType":1315,"score":1316,"id":1311},"\u002Fblog\u002Fpowering-tencent-billing-platform-with-apache-pulsar","Powering Tencent Billing Platform with Apache Pulsar","2019-10-22","blogs","Blog",1,{"path":1318,"title":1319,"date":1320,"image":1321,"link":-1,"collection":1314,"resourceType":1315,"score":1322,"id":1318},"\u002Fblog\u002Fa-practical-guide-to-enterprise-grade-security-in-apache-pulsar","A Practical Guide to Enterprise-Grade Security in Apache Pulsar","2023-02-16","\u002Fimgs\u002Fblogs\u002F640694ace2dfd06eaceb14a3_a-practical-guide-to-enterprise-grade-security-in-apache-pulsar.png",0.667,{"path":1324,"title":1325,"date":1326,"image":1327,"link":-1,"collection":1314,"resourceType":1315,"score":1322,"id":1324},"\u002Fblog\u002Fclient-optimization-how-tencent-maintains-apache-pulsar-clusters-100-billion-messages-daily","Client Optimization: How Tencent Maintains Apache Pulsar Clusters with over 100 Billion Messages Daily","2022-08-18","\u002Fimgs\u002Fblogs\u002F63c7c234f86b5b755c37354d_63b53070e64eae9f14d4b108_client-optimization-top-.jpeg",1775716409415]