Hubbry Logo
Oracle ExadataOracle ExadataMain
Open search
Oracle Exadata
Community hub
Oracle Exadata
logo
8 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Oracle Exadata
Oracle Exadata
Oracle Exadata
View on Wikipedia
from Wikipedia
Oracle Exadata
Original authorOracle Corporation
Initial releaseOctober 2008
Operating systemOracle Linux
PlatformExadata Database Machine, Exadata Database Service, Exadata Cloud@Customer
LicenseCommercial
Websitewww.oracle.com/exadata
Larry Ellison and Exadata (2009)

Oracle Exadata (Exadata[1]) is a computing system optimized for running Oracle Databases.

Exadata is a combined database machine and software platform that includes scale-out x86-64 compute and storage servers, RoCE networking, RDMA-addressable memory acceleration, NVMe flash, and specialized software.[2]

Exadata was introduced in 2008 for on-premises deployment, and since October 2015, via the Oracle Cloud as a subscription service, known as the Exadata Database Service on Dedicated Infrastructure,[3] and Exadata Database Service on Exascale Infrastructure.[4] Exadata Cloud@Customer[5] is a hybrid cloud (on-premises) deployment of Exadata Database Service.

Starting December, 2023, Exadata Database Service became available for Microsoft Azure, Google and AWS public clouds within the Oracle Database@Azure, Oracle Database@Google Cloud and Oracle Database@AWS multicloud partnerships.

Use cases

[edit]

Exadata is designed to run all Oracle database workloads, such as online transaction processing, data warehousing, analytics, and AI Vector processing, often with multiple consolidated databases running simultaneously.

Historically, specialized database machines were designed for a particular workload, such as Data Warehousing, and poor or unusable for other workloads, such as online transaction processing. Exadata specializes in mixed workloads sharing system resources with resource management features for prioritization, such as favoring workloads servicing interactive users over reporting and batch. Long running requests, characterized by Data Warehouses, reports, batch jobs and Analytics, are reported to run many times faster compared to a conventional, non-Exadata database server.[6][7]

Release history

[edit]
Exadata Release Primary Software Enhancements Primary Hardware Enhancements
Database@AWS Exadata Database Service available with AWS
X11M - Jan 2025 AI Vector search acceleration - up to 55% faster 25% faster compute core performance
Analytics scan throughput increase - 2.2x faster 33% greater server memory bandwidth
Transaction processing acceleration - 25% faster 11% faster storage core performance
Online transaction processing read latency acceleration - up to 21% faster (14 microseconds) PCIe 5 performance-optimized flash
Intelligent power management - reduce CPU cores, cap power consumption, optimize power utilization X11M-Z database and storage servers
Available on-premises, Oracle Cloud, Cloud@Customer and multicloud (Azure, Google Cloud, AWS) X11M-XT storage servers for less frequently accessed data. Supports Exascale.
Database@Google Cloud Exadata Database Service available with Google Cloud
Exadata Exascale

July, 2024

Fully elastic pay-per-use architecture. Users specify the cores and storage capacity needed, reducing entry-level infrastructure costs for Exadata Database Service and aligning costs with usage None
Large pools of shared compute and storage allow databases to quickly scale over time without concern for server-based size limitations or disruptive migrations
Rapid and efficient database snapshots and thin cloning
Database@Azure Exadata Database Service available with Microsoft Azure
X10M - June 2023 Exadata RDMA Memory (XRMEM) DRAM cache 3x increase in compute cores (96-core AMD EPYC)
Oracle Linux 8 and UEK 6 kernel updates 1.5x higher memory capacity
New In-Memory Columnar compression algorithm 2.5x faster DDR5 memory
Optimized Smart Scan for more complex queries 2.4x higher flash storage capacity (in all-flash storage)
Faster decryption and decompression 22% more disk storage capacity
X9M - Sept, 2021 Secure RDMA fabric isolation PCIe 4.0 dual-port active-active 100 Gb RoCE network
Smart Flash Log write-back 33% increase in compute cores
Storage Index and Columnar Cache persistence 33% increase in memory capacity
Faster decryption and decompression Algorithms 28% increase in disk capacity
Smart Scan performance optimizations 1.8x greater internal fabric bandwidth (PCIe 4.0)
1.8x greater flash bandwidth (PCIe 4.0)
X8M - Sept, 2019 RoCE: RDMA over Converged Ethernet Persistent Memory (PMEM) in storage
Persistent Memory Data Accelerator 100 Gbit/s internal fabric (2.5x increase)
Persistent Memory Commit Accelerator
KVM virtual machine support
X8 - April, 2019 AIDE: Advanced Intrusion Detection Environment Storage Server Extended (XT)
ML-based monitoring and auto-indexing 40% increase in disk capacity
Real-time updates of optimizer statistics 60% increase in storage processor cores
X7 - Oct, 2017 In-memory database in flash storage 2x increase in flash capacity
DRAM cache in storage 25% increase in disk capacity
Large-scale storage software updates 25 Gbit/s data center Ethernet support
Exadata Cloud@Customer Exadata Cloud Service on-premises
X6 - April, 2016 Exafusion direct-to-wire online transaction processing protocol 2x increase in flash capacity
Smart Fusion Block Transfer 10% increase in compute cores
Smart Flash Log 2x increase in memory capacity
Exadata Database Service Exadata on Oracle Cloud Infrastructure (OCI)
X5 - Dec, 2014 In-memory database fault tolerance 2x increase in flash & disk capacity
Database snapshots Elastic configurations
Xen virtual machine support All-flash storage server option
NVMe flash protocol support 50% increase in compute cores
IPv6 support 50% increase in memory capacity
X4 - Nov, 2013 Network Resource Management 2x increase in flash capacity
I/O latency capping 2x increase in memory capacity
Capacity-on-Demand licensing 50% increase in compute cores
Active/Active InfiniBand (2x increase) 33% increase in disk capacity
X3 - Sept, 2013 Smart Flash Cache write-back Eighth-Rack configuration
Improved management of slow disks/flash 4x increase in flash capacity
Sub-second brownout after storage failure 33% increase in compute cores
Simplified disk replacement 75% increase in memory capacity
Bypass predictive disk failure 2x increase in data center bandwidth
X2 - Sept, 2010 Smart Flash Log 8-socket (X2-8) configuration
Auto Service Request Storage Expansion Rack
Secure Erase of storage Hardware-based decryption
Platinum Services 50% increase in compute cores
2x increase in memory capacity
50% increase in disk capacity
8x increase in data center bandwidth
v2 - Sept, 2009 Storage Indexes Flash storage
Database-aware Smart Flash Cache Quarter-Rack configuration
Hybrid Columnar Compression 2x increase in memory & disk capacity
3x increase in data center bandwidth
40 Gbit/s internal fabric (2x increase)
v1 - Sept, 2008 Oracle Enterprise Linux Scale-out 4-socket compute servers
Smart Scan (storage offload) Scale-out 4-socket storage servers
IORM (I/O Resource Manager) 20 Gbit/s internal fabric (InfiniBand)
Join filtering (Bloom filters) 1 Terabyte disks
Incremental backup filtering 1 Gbit/s data center network (Ethernet)
Smart file creation

Support policy

[edit]

As the platform has been around since 2008, Oracle has published information related to the end-of-support for older Exadata generations. In Oracle's published document titled Oracle Hardware and Systems Support Policies,[8] they mention "After five years from last ship date, replacement parts may not be available and/or the response times for sending replacement parts may be delayed." To look up the "last ship date" of a particular Oracle Exadata generation, Oracle published a document titled Oracle Exadata - A guide for decision makers.[9]

Each generation of the Oracle Zero Data Loss Recovery Appliance shares components with similar generations of Exadata.

References

[edit]
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

Oracle Exadata

View on Grokipedia
from Grokipedia
Oracle Exadata is an integrated database platform that combines hardware and software engineered for workloads, including , , AI vector search, and mixed environments. Introduced in 2008 as the first product in Oracle's family of Engineered Systems, Exadata was designed to address the limitations of general-purpose hardware by providing a tightly integrated system tailored specifically for , enabling faster query processing and reduced through database-aware storage and networking optimizations. At its core, Exadata consists of scale-out racks featuring high-performance database servers, intelligent storage servers with flash and disk drives, high-speed or Ethernet networking, and the Exadata , which offloads data-intensive operations from the database to storage for improved and OLTP performance. It supports flexible deployment models, including on-premises in enterprise data centers via the Exadata Database Machine, in Infrastructure as Exadata Database Service, on-premises cloud via Exadata Cloud@Customer behind customer firewalls, and multicloud integrations with AWS, Azure, and Google Cloud, allowing organizations to run 19c alongside the latest AI Database 26ai. Key benefits include automated management through Autonomous Database capabilities, independent scaling of compute and storage resources, and enhanced security features like data encryption and zero-trust , which reduce costs via consolidation and eliminate manual tuning for routine tasks. The platform's evolution continues with the 13th generation Exadata X11M, announced on January 7, 2025, incorporating Exadata System Software 25ai for advanced AI workloads and supporting quarterly updates to ensure compatibility with emerging database innovations.

Overview

Definition and Purpose

Oracle Exadata is an engineered system that combines a database machine with specialized software, utilizing scale-out compute and storage servers interconnected via (RoCE) networking and equipped with NVMe flash storage. This integrated platform is designed specifically for deploying in enterprise environments, providing a preconfigured, full-stack solution that optimizes at both the hardware and software levels. The primary purpose of Oracle Exadata is to achieve extreme performance, scalability, and efficiency for workloads, encompassing (OLTP), , and (AI) applications. It accomplishes this by offloading significant portions of data processing from database servers to intelligent storage servers, reducing data movement across the network and minimizing CPU overhead on the compute nodes. This offloading mechanism, known as Exadata Smart Scan, enables storage cells to perform predicate filtering, column projection, and other SQL operations directly on data blocks before transmission, thereby accelerating query execution and supporting massive-scale deployments. Introduced in September 2008 through a partnership between and , Oracle Exadata has evolved into a fully Oracle-engineered , with subsequent iterations incorporating advancements in hardware and software to enhance its capabilities. This progression has solidified its role as a dedicated platform tailored exclusively for , allowing for deep, database-aware optimizations that are infeasible on generic hardware configurations.

Key Benefits

Oracle Exadata provides significant performance advantages for workloads through its integrated hardware and software design, enabling up to 10x faster query processing compared to traditional systems. This acceleration is primarily achieved via Smart Scan and SQL offloading technologies, which process data directly on storage servers, reducing data transfer to the database servers and minimizing latency for mixed OLTP and workloads. Benchmarks demonstrate reduced response times, with OLTP latency as low as 14 microseconds in recent models such as Exadata X11M (as of 2025), representing up to 25% improvement over the prior generation and over 17x faster than traditional systems. Cost efficiencies are a core benefit of Exadata, driven by infrastructure consolidation that allows hundreds of to run on a single system, reducing the need for multiple servers and lowering (TCO). This consolidation simplifies and optimizes resource utilization, enabling organizations to handle the same workloads with fewer physical components and less operational overhead. Studies and implementations show TCO reductions of up to 48% through such efficiencies, particularly when combined with that streamlines deployment and maintenance. Exadata's supports linear growth for massive , handling up to exabytes of data while maintaining , making it suitable for consolidating hundreds of on one platform. Its scale-out allows seamless expansion by adding racks without disrupting operations, accommodating diverse workloads from terabytes to petabytes efficiently. Automation features in Exadata, including built-in tools for patching, tuning, and scaling, minimize manual DBA intervention and enhance operational efficiency. These autonomous capabilities use to proactively manage resources, apply updates without downtime, and optimize performance automatically, reducing administrative costs and errors. For mission-critical applications, Exadata incorporates robust security measures such as (TDE) for data at rest and Secure RDMA Fabric for network isolation, ensuring compliance and protection against unauthorized access. With Exadata System Software 25ai in the X11M generation, it further supports advanced AI workloads including vector search.

History

Origins and Development

Oracle Exadata originated from a joint engineering project between and (HP) initiated in 2008 to tackle persistent database I/O bottlenecks in high-performance environments. The collaboration focused on integrating software with specialized storage hardware, enabling offloading of data processing tasks from the database servers to intelligent storage cells, which dramatically reduced latency and improved query . Initial prototypes were developed and rigorously tested within Oracle's internal laboratories, validating the concept of hardware-software co-design for database optimization before public announcement at Oracle OpenWorld in September 2008. This marked the debut of the HP Oracle Database Machine, the first iteration of Exadata (V1), as Oracle's inaugural engineered system tailored for data warehousing and analytics workloads. A pivotal shift occurred in 2009 following Oracle's announced acquisition of Sun Microsystems in April of that year, which enabled the transition from HP hardware to Sun-based platforms for subsequent development. The acquisition, completed on January 27, 2010, granted full control over hardware engineering, allowing deeper integration of its database expertise with Sun's server and storage technologies. This evolution culminated in Exadata V2, announced at Oracle OpenWorld 2009, which expanded capabilities to support (OLTP) alongside analytics, solidifying Oracle's proprietary engineered systems approach. The development philosophy underpinning Exadata emphasized "engineered systems," where hardware and software are co-designed specifically for to deliver optimized performance, scalability, and reliability beyond commodity configurations. This strategy drew directly from Oracle's decades of database innovation, prioritizing seamless integration to eliminate traditional silos between compute, storage, and networking. Early adoption began immediately after the launch, with initial deployments targeted at high-end enterprise customers in sectors such as banking and , where the storage offloading proved essential for handling massive volumes and real-time demands. These pioneering implementations validated Exadata's value in mission-critical environments, paving the way for broader among Global 100 companies.

Major Releases

The first commercial release of Oracle Exadata, designated as version V1, occurred in September 2008 and was based on hardware, featuring interconnects for high-speed data transfer in data warehousing environments. In September 2009, version V2 shifted to hardware, introducing quarter-rack configurations, flash caching for (OLTP) workloads, and hybrid columnar compression to enhance storage efficiency. The X2 generation launched in September 2010, incorporating networking upgrades, increased core counts up to 96 per rack, and smart flash logging for improved I/O performance. Subsequent releases built on this foundation: X3 in September 2012 added 40% more flash capacity and write-back flash cache; X4 in December 2013 introduced all-flash storage options and network resource management; and X5 in 2015 integrated NVMe SSDs for faster access, alongside initial cloud service compatibility. From 2016 to 2019, the X6, X7, and X8 generations advanced progressively with doubled flash capacities, higher core densities, PCIe 4.0 support, and features like hot-pluggable flash and 25 Gb/sec Ethernet in X7, culminating in machine learning-based monitoring in X8. More recent innovations include the X10M in 2023, which delivered three times the compute density using processors and RoCE v2 networking; and the X11M in January 2025, emphasizing AI vector search acceleration and a compact 2RU form factor for enhanced efficiency. Oracle maintains a release cadence of new hardware generations every 12 to 24 months, supplemented by quarterly updates to Exadata for ongoing performance and security enhancements. Over time, Exadata's evolution has shifted from a primary on-premises focus to greater cloud integration, beginning with Exadata Cloud Service in 2015. End-of-support timelines vary by generation, with premier support typically lasting five years post-last ship date; for example, X8 reaches end-of-support in December 2025, while parts availability extends five years after the last ship date for eligible systems.

Architecture

Hardware Components

Oracle Exadata systems are built on a scale-out comprising high-performance compute servers, intelligent storage servers, and integrated networking fabric, all optimized for database workloads. The hardware has evolved across generations to incorporate advanced processors, technologies, and storage media, enabling configurations from compact quarter racks to large-scale multi-rack clusters. Compute servers, known as Database Machine (DM) servers, serve as the primary processing nodes for Oracle Database instances. In the latest Exadata X11M generation, each DM server features two 96-core AMD EPYC 9J25 processors operating at 2.6 GHz base frequency (boosting up to 4.5 GHz), providing substantial parallel processing capability for OLTP and analytics tasks. Memory capacity reaches up to 3 TB per node using 6400 MT/s DDR5 DIMMs, supporting in-memory database operations and large-scale caching. System storage includes two 3.84 TB NVMe SSDs, expandable to four, for local OS and database files. These servers utilize PCIe Gen5 interfaces for enhanced I/O bandwidth, including RDMA Network Fabric adapters delivering 200 Gb/s combined throughput. Additionally, a new Database Server-Z variant offers a single 32-core x86 processor with up to 1.152 TB DDR5 memory in a more compact form, targeted at lighter workloads. Storage servers in Exadata handle data persistence, offloading, and intelligent caching, available in hybrid or all-flash variants. Exadata X11M High Capacity (HC) storage servers include two 32-core AMD EPYC 9J15 processors at 2.95 GHz (up to 4.4 GHz) and 1.5 TB of DDR5 RAM, with 12 x 22 TB HDDs for bulk storage totaling 264 TB raw capacity per server, augmented by 27.2 TB of performance-optimized flash for Smart Flash Cache and Log. In contrast, Extreme Flash (EF) configurations replace HDDs with eight NVMe flash drives—four 30.72 TB capacity-optimized and four 6.8 TB performance-optimized—yielding 150.08 TB raw flash per server for low-latency, high-IOPS applications. All storage servers support PCIe Gen5 and incorporate Exadata RDMA Memory (XRMEM) cache from system DRAM for accelerated data access. These servers operate in 2U form factors, consistent with recent generations' shift toward denser, energy-efficient designs. Rack configurations provide flexible deployment options, starting from a base quarter rack with two compute servers and three storage servers, scaling to half racks (four compute and six storage) or full racks (eight compute and 12-19 storage, depending on elastic setup). Elastic configurations allow customization within a single rack, such as two compute with 17 storage for capacity-focused setups or up to 15 compute with three storage for compute-intensive needs, all housed in standard 42U cabinets with integrated power distribution units rated up to 22.5 kW maximum draw. Cooling relies on front-to-back airflow with redundant fans, supporting high-density operations in data centers. Multi-rack clusters interconnect via RDMA fabric for seamless scaling. Over generations, Exadata hardware has progressed from earlier 1U and 2U form factors to standardized 2U in X11M, emphasizing higher core counts, DDR5 , and PCIe Gen5 for bandwidth exceeding 100 Gb/s per . This evolution enables capacity scaling to exascale levels, with multi-rack Exascale configurations supporting over 100 PB of raw storage through software-defined pooling across hundreds of storage servers.

Software Components

Oracle Exadata's software stack integrates tightly with its hardware to deliver optimized database performance, leveraging specialized components for management, offloading, and operations. The core software includes the , which runs on Exadata systems with versions such as 19c, 21c, 23ai, and 26ai, enhanced by Exadata-specific patches that enable advanced offloading capabilities like Smart Scan and Hybrid Columnar Compression. These patches ensure seamless integration between the and Exadata's storage cells, allowing for efficient data processing directly on storage hardware. The Exadata System Software 25ai serves as the foundational layer for system management, utilizing an image-based deployment model that simplifies updates and maintenance across database and storage servers. As of 2025, versions such as 25.1 and 25.2 provide key tools including CellCLI, a for configuring and monitoring Exadata storage cells, and ExaCLI, which enables remote execution of commands for system-wide operations. This software manages critical functions like I/O resource allocation and performance metrics collection, ensuring high efficiency in data-intensive environments. Quarterly updates, such as release 25.1.10 in October 2025, address bugs, security vulnerabilities, and introduce features like AI Smart Scan enhancements for improved query processing with workloads. At the operating system level, Exadata employs Oracle Linux equipped with the Unbreakable Enterprise Kernel (UEK), a customized kernel optimized for enterprise workloads and Exadata's architecture. This OS supports Remote Direct Memory Access (RDMA) for low-latency networking between servers and storage, reducing CPU overhead in data transfers. Additionally, Oracle Ksplice enables zero-downtime kernel updates by applying patches without rebooting, maintaining continuous availability for mission-critical databases. Management tools further streamline Exadata operations, with the Exadata Deployment Assistant (EDA) facilitating initial system configuration by generating scripts for network setup, storage allocation, and software installation based on user-specified parameters. For ongoing monitoring and administration, (OEM) provides a centralized interface to track performance, alerts, and resource utilization across the Exadata rack, integrating with Exadata-specific plugins for detailed diagnostics. These tools align with the quarterly software update cadence to incorporate the latest security and feature enhancements.

Networking and Integration

Oracle Exadata employs a high-performance internal networking fabric to interconnect compute and storage servers, enabling efficient data transfer and scalability. Starting with the X8M generation, Exadata transitioned from to (RoCE) as the primary internal fabric, providing low-latency communication across system components. In X10M and later models, this fabric utilizes dual-port PCIe Gen 5 network interface cards (NICs) supporting 2 x 100 Gb/sec active-active RoCE connections, delivering a total throughput of 200 Gb/sec per server while minimizing CPU overhead through . The RoCE implementation incorporates features like Priority Flow Control (PFC) and (ECN) to ensure zero and prioritize critical database traffic, achieving latencies under 17 microseconds for memory-intensive operations. Client access to Exadata systems occurs via dedicated Ethernet interfaces, supporting speeds of 10 Gb/sec, 25 Gb/sec, or 100 Gb/sec to connect applications and external networks. These interfaces, typically configured as bonded pairs for , allow seamless integration with enterprise LANs and provide high-bandwidth entry points for transactional and analytical workloads without compromising internal fabric performance. A key integration aspect of the Exadata networking fabric is its RDMA-aware design, which facilitates direct transfers between database servers and storage servers, bypassing traditional I/O stacks and reducing latency for data-intensive queries. This capability, enhanced by Exafusion protocols, enables offloaded database operations directly over the fabric, such as Smart Scan and columnar caching, ensuring efficient resource utilization across the system. For scalability, Exadata supports multi-rack clustering through a leaf-spine (fat-tree) , where leaf switches connect to servers within racks and spine switches interconnect racks for non-blocking communication. Configurations can scale up to 14 racks using RoCE without requiring external switches, allowing linear increases in I/O throughput—for instance, four racks provide four times the bandwidth of a single rack—while maintaining consistent performance for large deployments. In Exascale environments, this architecture extends to manage fleets of storage servers over the RDMA fabric, supporting cloud-scale elasticity for AI and . Networking security in Exadata includes integrated support and host-based firewalls to isolate traffic and enforce policies. RoCE enable secure fabric isolation, preventing inter-VM cluster visibility and ensuring encrypted, segmented communications across the internal fabric. Additionally, firewalls on database and storage servers provide granular control over inbound and outbound traffic, complementing segmentation for compliance with enterprise security standards.

Features

Database Offloading Technologies

Oracle Exadata employs database offloading technologies to shift data-intensive from database servers to intelligent storage servers, minimizing network and alleviating CPU burdens on the database tier. This approach enables the storage layer to perform operations such as filtering and aggregation directly on disk , returning only pertinent results to the database servers. By leveraging these mechanisms, Exadata optimizes query for both transactional and analytical workloads, ensuring scalable efficiency in large-scale environments. A cornerstone of this offloading is Smart Scan, which processes table scans and index scans at the storage level by evaluating database predicates and projecting specific columns. During a Smart Scan, the storage servers receive query directives from the database and scan data blocks in parallel, applying filters to eliminate irrelevant rows before transmission, thereby scanning only the necessary portions of data blocks. This technology supports a wide range of SQL operations, including joins, aggregations, and sorting, executed efficiently on storage hardware. Complementing Smart Scan are storage indexes, in-memory structures maintained on each Exadata storage server that track minimum and maximum values for columns across predefined storage regions, typically 1 MB in size. These indexes enable rapid query routing by allowing storage servers to prune regions unlikely to contain qualifying data, thus avoiding full table scans and reducing I/O operations. Storage indexes are automatically created and updated as data is modified, supporting up to 24 columns per index in recent releases and enhancing selectivity for range-based predicates. Exadata Smart Flash Cache further bolsters offloading by intelligently caching hot data on high-performance NVMe flash storage within the storage servers. Operating as a write-through cache for decision support systems and a write-back cache for , it prioritizes frequently accessed blocks to accelerate read operations while integrating seamlessly with Smart Scan to process cached data . This caching mechanism dynamically evicts less useful data and supports manual population for specific workloads, ensuring low-latency access to critical datasets without burdening database servers. Offload processing is facilitated by the Cell Offload Server (COS), a sub-process of the Cell Server (CELLSRV) on storage servers that handles version-specific offload requests from instances. COS enables SQL execution on storage cells by interpreting and processing directives for operations like predicate evaluation and , reducing the volume of data sent over the network. This integration allows for hybrid columnar compression awareness during offloading, briefly enhancing efficiency in analytical queries as detailed in related storage techniques. In Exadata System Software 25ai, introduced in 2025, AI Smart Scan enhancements accelerate AI vector search queries in 26ai by offloading vector operations to storage servers. This provides up to 8x faster performance for INT8 vector formats and 32x for BINARY formats, enabling efficient processing of AI workloads at scale. Oracle Exadata supports OLAP (Online Analytical Processing) workloads for complex analysis, data warehousing, and queries on large historical data volumes. This capability is enabled through Smart Scan and offloading processing to storage servers, which optimize massive scans and analytical queries by performing filtering, aggregation, and other operations at the storage level, delivering up to 31 TB/sec of scan throughput for data warehouse databases as large as 40 petabytes. The cumulative impact of these offloading technologies significantly enhances performance, particularly for analytical queries, by reducing data transfer between storage and database servers by over 90% in many scenarios through selective filtering and projection. This results in accelerated query execution times, elimination of I/O bottlenecks, and increased overall system throughput, enabling Exadata to handle petabyte-scale datasets with sub-second response times in consolidated environments.

Storage and Compression

Oracle Exadata employs Hybrid Columnar Compression (HCC) as its primary storage optimization technique, enabling significant reductions in footprint for analytic workloads. HCC organizes into compression units where rows are stored in a columnar format, grouping similar values together to facilitate efficient encoding. This method is particularly effective for tables loaded via direct-path inserts, achieving compression ratios of up to 10x on average, with ranges from 5x to 20x depending on characteristics and compression mode. HCC is especially beneficial for OLAP workloads, supporting efficient storage and decompression during queries on large historical data volumes in data warehousing environments. The compression process leverages dictionary encoding to replace repeated values with numeric references and to compact sequences of identical values, all within fixed-size compression units of approximately 1 MB. Warehouse mode prioritizes query performance with moderate compression (up to 10x for high settings), while archive mode maximizes space savings (up to 15x) for infrequently accessed data. These techniques maintain compatibility with Exadata's query offload capabilities, where decompression occurs transparently on storage servers without burdening database CPUs. Exadata's storage architecture incorporates automated tiering across multiple media types to optimize capacity and access speed. Cold resides on high-capacity hard disk drives (HDDs) in High Capacity (HC) storage servers, providing economical bulk storage of up to 264 TB raw per server. Warm is placed on solid-state drives (SSDs) in Extreme Flash (EF) configurations, offering up to 122.88 TB raw capacity per server for balanced performance and density. Hot is accelerated via the Smart Flash Cache, utilizing NVMe flash drives (up to 27.2 TB per server) and Exadata RDMA (XRMEM) for sub-millisecond latencies. The Exadata dynamically manages movement between these tiers based on access patterns, ensuring frequently used blocks are promoted to faster layers while retaining less active on slower, higher-capacity media. This tiered approach, combined with HCC, delivers substantial effective capacity in production deployments. For instance, a full Exadata X10M rack with HC storage servers provides up to 4.2 PB raw disk capacity, expanding to over 42 PB usable with 10x average compression, scalable further through expansion racks to exceed 50 PB in larger configurations. Such optimizations reduce physical storage requirements and operational costs without compromising accessibility.

High Availability and Security

Oracle Exadata incorporates multiple layers of redundancy to ensure across its hardware components. Each storage server features dual controllers for automatic in case of controller failure, with Exadata System Software enabling seamless redirection of I/O operations to maintain data availability. Power supplies operate in an configuration, providing redundant power distribution units (PDUs) and hot-swappable units (PSUs) to prevent single points of failure. Additionally, Oracle Automatic Storage Management (ASM) high redundancy maintains a primary copy and two mirrored copies of data, automatically repairing corruptions and rebalancing data upon disk or flash failures without downtime. High availability is enhanced through tight integration with Oracle Real Application Clusters (RAC), which supports cache fusion for shared data access across nodes. In Exadata, cache fusion operates over the (RoCE) network fabric, enabling low-latency block transfers and rapid failure detection in under 2 seconds via Instant Failure Detection (IFD). For disaster recovery, provides zero data loss protection using synchronous redo transport, replicating changes to up to 30 standby databases at rates up to 500 MB/sec for OLTP workloads, with automatic failover capabilities to minimize downtime. These mechanisms achieve near-zero brownout during storage failures, with recovery times typically in seconds to minutes. Backup and recovery processes are optimized for Exadata using Recovery Manager (RMAN), which supports high-performance s to Oracle ZFS Storage Appliances. RMAN leverages Exadata's offloading capabilities to create sets or copies with up to 3,000 concurrent threads, distributing I/O across multiple paths for efficiency while using block-change tracking to accelerate incremental s. This integration ensures rapid recovery without impacting production performance, complementing Exadata's resilient storage architecture. Security in Exadata emphasizes data protection at rest and in transit, alongside controlled access. (TDE), part of Oracle Advanced Security, encrypts tablespaces, columns, temporary data, redo logs, backups, and even the Exadata Smart Flash Cache, using FIPS 140-2 compliant modules without requiring application changes. Network traffic is secured via native encryption or SSL/TLS protocols, including hardware-accelerated AES-NI for Oracle Net Services, JDBC, and communications to Data Guard standbys, while syslog transfers use certificate-based TLS starting with Exadata 19.3.0. is enforced through Exadata , where administrators create roles with fine-grained privileges using CellCLI commands and assign them to users authenticated via the Management Server with hashing. The ExaCLI utility facilitates secure remote management of storage servers over and APIs, requiring specified user roles for command execution and supporting between database and storage administration.

Deployment Options

On-Premises and Cloud@Customer

Oracle Exadata supports on-premises deployments through the Exadata Database Machine, where customers take full ownership of X-series racks such as the X11M model, enabling organizations to operate a private database tailored to their needs. These systems are scalable, starting from a quarter-rack configuration that includes two database servers and three storage servers, and expanding to full racks or multi-rack setups to accommodate growing workloads without downtime. provides professional installation services to ensure proper setup, allowing customers to focus on application integration rather than hardware assembly. Exadata Cloud@Customer extends this capability by deploying the full Exadata Database Service directly in the customer's , with Oracle handling all management and operations through Oracle Cloud Infrastructure (OCI). This hybrid model maintains the performance and features of public cloud Exadata while keeping data on-premises to meet regulatory and requirements. As of October 2025, Generation 2 updates for Exadata Cloud@Customer include an optional 100Gbps backup network interface card (NIC), which replaces the standard 25Gbps NIC and includes dual 100Gbps cards with transceivers to accelerate backups and reduce impact on production systems. The setup process for both on-premises and Cloud@Customer deployments begins with rack delivery to the customer site, followed by physical cabling for power, networking, and internal connections as specified in the installation guide. System imaging is then performed using the Exadata Deployment Assistant (OEDA), a tool that generates configuration files and automates the deployment of operating systems, Oracle Grid Infrastructure, and databases across the rack's components. Power requirements vary by configuration but typically demand a reliable source, such as up to 21.3 kVA three-phase for a full high-capacity rack, with two power distribution units (PDUs) per rack to ensure and prevent outages. Management of on-premises Exadata systems involves local access via the Integrated Lights Out Manager (ILOM), an embedded service processor that enables monitoring, firmware updates, and hardware diagnostics for all servers and storage units. For Cloud@Customer, manages these aspects remotely through OCI, while still providing integration with Support for proactive issue resolution and telemetry data sharing. These deployment options offer key advantages, including to comply with local regulations and low-latency access ideal for integrating with legacy applications in existing data centers. Unlike public cloud alternatives, they allow customers to retain control over their physical environment while benefiting from Exadata's engineered optimizations.

Cloud and Multicloud

Oracle Exadata Database Service on Oracle Cloud Infrastructure (OCI) provides a fully managed deployment option, allowing users to provision Exadata infrastructure in the cloud without managing underlying hardware. Available shapes range from quarter-rack equivalents for smaller workloads to full-rack configurations for enterprise-scale demands, enabling flexible resource allocation based on performance needs. This service integrates seamlessly with Oracle Autonomous Database, supporting automated provisioning and management of autonomous workloads on Exadata, which enhances scalability and reduces administrative overhead. Auto-scaling capabilities allow dynamic adjustment of compute resources, such as OCPUs, to handle varying loads without downtime, optimizing costs and performance for mission-critical applications. In multicloud environments, Oracle extends Exadata support to third-party clouds, starting with Oracle Database@Azure in December 2023, which deploys dedicated Exadata infrastructure within data centers for low-latency access to Oracle databases. Similar offerings include Oracle Database@Google Cloud, launched in 2024, and Oracle Database@AWS, generally available in July 2025, both providing Exadata Database Service on dedicated infrastructure to facilitate hybrid architectures and avoid . The Globally Distributed Exadata Database on Exascale Infrastructure, generally available in August 2025, enables cross-region and sharding across multiple cloud providers, supporting high-availability setups with sub-second replication for global applications. Key features of cloud and multicloud Exadata deployments include pay-as-you-go models, which charge based on actual such as OCPUs and storage, offering cost predictability without upfront commitments. 23ai introduces AI Vector Search capabilities, allowing efficient indexing and querying of vector embeddings for AI-driven applications directly on Exadata in the . Backups can be configured to OCI for durable, scalable off-site protection, with automated policies ensuring compliance and recovery readiness. For migrations, Oracle Zero Downtime Migration (ZDM) tool supports lift-and-shift strategies, enabling physical or logical online transfers from on-premises Exadata to instances with minimal interruption. Recent 2025 updates enhance cloud compatibility, including support for Exadata X11M hardware in OCI deployments running Exadata System Software version 25.1, which introduces optimizations for AI workloads and improved power efficiency.

Use Cases

Transaction Processing

Oracle Exadata is optimized for high-volume (OLTP) workloads, enabling real-time transaction handling through its integrated hardware and . Key optimizations include the Exadata Smart Flash Cache, which serves as a low-latency read-and-write cache, providing OLTP I/O latencies as low as 19 microseconds for reads and supporting write-back caching to accelerate commit operations for write-intensive applications. This flash technology prioritizes frequently accessed data blocks, reducing disk I/O and ensuring sub-second query response times even under heavy loads. Additionally, while the In-Memory Column Store primarily enhances analytical scans, its dual-format storage indirectly benefits OLTP by minimizing index maintenance overhead and enabling efficient mixed workloads. Scalability in is achieved via Real Application Clusters (RAC) on Exadata, which supports horizontal scaling across multiple nodes with RDMA-enabled interconnects for low-latency messaging and improved throughput. Exadata supports high transaction rates with linear scalability across racks for hyperscale environments. Auto-sharding further distributes workloads dynamically, maintaining consistency and performance in distributed setups. For instance, in financial trading systems, Exadata's ensures sub-millisecond latencies for order matching and settlement. Real-world applications include order processing, where Exadata handles peak transaction surges, and systems for real-time account updates. Integration with and APEX enhances enterprise applications by leveraging Exadata's low-latency storage for faster form submissions and dynamic reporting, streamlining development and deployment. Overall, these features enable Exadata to process mixed OLTP/OLAP environments efficiently, with write-back flash reducing commit times by offloading redo log writes to high-speed persistent memory.

Analytics and AI

Oracle Exadata supports OLAP (Online Analytical Processing) for complex analysis, data warehousing, and queries on large historical data volumes using Smart Scan, Hybrid Columnar Compression (HCC), and offloading processing to storage for optimized massive scans and analytical queries. Oracle Exadata excels in data warehousing by leveraging parallel query execution and Smart Scan technology, which offloads filtering and aggregation directly to the storage servers, enabling efficient of massive datasets. This approach allows for terabyte-scale scans to complete in minutes by minimizing movement across the network and reducing CPU load, as demonstrated in benchmarks where Exadata systems process petabyte-scale rapidly through distributed storage . In analytics workloads, such as reporting and ad-hoc queries, features like Exadata Hybrid Columnar Compression accelerate scans by 10-15 times compared to hardware setups by enabling without full decompression, with Smart Scan optimizing the integration for tasks. For AI and support, Exadata integrates 23ai's AI Vector Search, which enables searches on vector embeddings stored alongside relational , facilitating applications like retrieval-augmented generation (RAG) pipelines for generative AI. This capability allows developers to query vectors using SQL for contextual in enterprise , with Exadata's offloading extending to vector operations for up to 30 times faster AI workloads on the X11M platform through intelligent storage acceleration. Additionally, GPU integration via partnerships in Exadata environments, such as Compute Cloud@Customer combined with X11M, enhances vector search and inference for large language models, supporting consolidated processing for AI tasks. Practical examples include ad-hoc in retail, where Exadata enables rapid querying of customer transaction data for insights like , as seen in deployments handling unpredictable growth in sales . For generative AI model training, Exadata consolidates datasets on a single platform, allowing efficient preparation and fine-tuning of models on historical data. This consolidation supports resource isolation through features like database resource manager and pluggable databases, ensuring and AI tasks do not interfere with other operations while maximizing hardware utilization. In 2025, Exadata's Exascale Infrastructure advances agentic AI by providing globally distributed databases with extreme availability and performance, offloading vector queries to storage for scalable semantic processing in workflows. This setup supports high-throughput AI applications, including vector databases handling up to billions of embeddings per instance, aligning with the demands of agentic systems that require real-time, low-latency interactions across petabyte-scale .

Support and Maintenance

Support Policies

Oracle's Lifetime Support Policy for Exadata encompasses three phases designed to provide ongoing technical assistance as long as products remain licensed. Premier Support lasts five years from the general availability date of each release, offering major updates, new certifications, security alerts, and full technical support. Extended Support provides an optional three-year extension beyond Premier, including continued updates and fixes but limited to existing certifications without support for new products, available for an additional fee. Sustaining Support follows indefinitely, granting access to existing patches, bug fixes from prior phases, and technical assistance via My Oracle Support, though no new updates or certifications are issued. Hardware support for Exadata systems includes a standard five-year warranty covering onsite service and replacement parts, with Premier Support for Systems extending at least five years from the last ship date. Parts replacement is provided using new or equivalent quality components during this period, after which availability may diminish and response times extend. For example, support for Exadata X8 systems concluded on October 20, 2025, marking the end of active hardware servicing for that generation. Software maintenance under Exadata support involves quarterly releases of Exadata System Software, such as version 25ai (25.1) released in December 2024, which deliver feature enhancements, bug fixes, and updates. Additionally, monthly patches are applied via Ksplice, enabling zero-downtime updates for critical kernel vulnerabilities and other components. Support entitlements for Exadata customers include 24/7 access to specialized engineers for and resolution, with remote diagnostics facilitated by ExaCHK, a non-intrusive tool that automates system monitoring and compliance verification across hardware and software stacks. On-site response agreements (SLAs) target Severity 1 issues within two hours for locations up to 25 miles from an service center, extending to four hours for 26-49 miles, ensuring rapid hardware intervention. For systems reaching end-of-life (EOL), third-party providers offer extended support alternatives, including break-fix services, monitoring, patching assistance, and hardware coverage.

Upgrades and Lifecycle Management

Oracle Exadata systems support upgrade paths that encompass both software and hardware components to ensure ongoing performance and compatibility. Software upgrades are facilitated through rolling updates using the patchmgr utility, which orchestrates updates across storage servers, database servers, and RDMA Network Fabric switches in a sequential manner to minimize . This approach allows one component at a time to be taken offline, updated with operating system, firmware, and Exadata System Software patches, and then brought back online before proceeding to the next. For hardware refreshes, Oracle provides options for in-place expansions and updates, though specific compute and storage swaps require coordination with Oracle support to maintain system integrity. Key tools streamline the upgrade process and facilitate migrations. The Exadata patchmgr utility serves as the primary software patching tool, automating the staging, application, and rollback of updates while supporting both rolling and non-rolling modes for flexibility. For cloud shifts, Oracle Zero Downtime Migration (ZDM) enables seamless database transfers from on-premises Exadata to cloud environments with minimal interruption, using physical or logical methods to replicate data online. Lifecycle management for Exadata spans planning, operation, and decommissioning stages. In the planning phase, capacity forecasting utilizes Oracle Cloud Infrastructure Operations Insights to analyze trends in CPU, storage, memory, and I/O utilization, enabling proactive scaling decisions based on historical data and machine learning predictions. During operation, monitoring is conducted via Oracle Exachk (EXAchk), a non-intrusive health check framework that scans hardware, software, and configuration across the Exadata stack to identify potential issues and ensure compliance. Decommissioning involves data migration using tools like ZDM to transfer workloads to newer systems or cloud instances, followed by secure erase procedures on storage servers to wipe persistent memory and drives, preventing data remnants. Best practices emphasize regular maintenance to sustain reliability and security. A quarterly patching schedule is recommended, aligning with Oracle's infrastructure maintenance windows to apply critical updates during low-impact periods, often incorporating backups beforehand to enable rollbacks if needed. Compatibility matrices confirm support for 19c, Oracle AI Database 23ai (minimum Database 23.5), and Oracle AI Database 26ai (October 2025 release update) on Exadata X11M systems, requiring minimum Exadata 25ai (25.1.0). In 2025, Exadata X11M introduces upgrade enhancements tailored for AI workloads, including AI Smart Scan optimizations that offload vector search operations to storage for low-latency processing of INT8 and binary vectors. Exadata Cloud Infrastructure X8 SKUs reached end-of-life on October 20, 2025, necessitating migration to newer models like X11M via ZDM to avoid support discontinuation.

References

  1. https://www.oracle.com/engineered-systems/exadata/
  2. https://www.oracle.com/database/technologies/exadata/exadata-dmg/
  3. https://www.oracle.com/webfolder/s/assets/infographics/10-years-of-exadata/index.html
  4. https://www.oracle.com/database/technologies/exadata/architecture/
  5. https://www.oracle.com/database/technologies/exadata/hardware/
  6. https://www.oracle.com/engineered-systems/exadata/database-service/
  7. https://www.oracle.com/engineered-systems/exadata/cloud-at-customer/
  8. https://www.oracle.com/engineered-systems/exadata/database-machine/
  9. https://www.oracle.com/database/technologies/exadata/
  10. https://www.oracle.com/database/technologies/exadata/hardware/rdmanetwork/
  11. https://www.oracle.com/a/ocom/docs/engineered-systems/exadata/exadb-d-x11m-ds.pdf
  12. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/offloading-data-search-and-retrieval-processing.html
  13. https://www.oracle.com/us/corporate/analystreports/industries/059641.pdf
  14. https://www.oracle.com/technetwork/server-storage/engineered-systems/exadata/dbmachine-x3-twp-1867467.pdf
  15. https://blogs.oracle.com/exadata/exadata-x11m
  16. https://www.oracle.com/news/announcement/oracle-introduces-exadata-x11m-platform-2025-01-07/
  17. https://www.oracle.com/docs/tech/exadata-for-consolidation.pdf
  18. https://www.oracle.com/a/ocom/docs/autonomous-tco-report.pdf
  19. https://www.oracle.com/a/ocom/docs/fastest-database-machine-exadata-cloud-customer.pdf
  20. https://docs.oracle.com/en/cloud/paas/autonomous-database/dedicated/adbaa/index.html
  21. https://www.oracle.com/autonomous-database/features/
  22. https://docs.oracle.com/en/cloud/paas/exadata-cloud/csexa/data-security.html
  23. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmsq/using-cryptographic-services.html
  24. https://www.oracle.com/corporate/pressrelease/oracle-buys-sun-042009.html
  25. https://docs.oracle.com/en/engineered-systems/
  26. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/software-release-availability.html
  27. https://www.oracle.com/support/lifetime-support/
  28. https://www.natrinsic.com/exadata
  29. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/oracle-exadata-database-server-x11m-hardware-components.html
  30. https://www.oracle.com/au/a/ocom/docs/engineered-systems/exadata/exadata-x11m-ds.pdf
  31. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/x11m-high-capacity-exadata-storage-server-components.html
  32. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/x11m-extreme-flash-exadata-storage-server-components.html
  33. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/exadata-database-machine-configuration.html
  34. https://www.oracle.com/database/technologies/exadata/exascale/
  35. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/exadata-software-overview.html
  36. https://support.oracle.com/knowledge/Oracle%20Database%20Products/2333222_1.html
  37. https://docs.oracle.com/en-us/iaas/exadata/doc/ecc-system-config-options.html
  38. https://www.oracle.com/news/announcement/ai-world-database-26ai-powers-the-ai-for-data-revolution-2025-10-14/
  39. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/oracle-exadata-storage-server-software-components.html
  40. https://blogs.oracle.com/exadata/post/software-2025-aug
  41. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/whats-new-oracle-exadata-system-software-release-25.1.html
  42. https://blogs.oracle.com/exadata/software-2025-oct
  43. https://www.oracle.com/a/ocom/docs/engineered-systems/exadata/exadata-x10m-ds.pdf
  44. https://blogs.oracle.com/exadata/post/maximize-database-performance-with-oracle-exadata-and-oracle-linux
  45. https://docs.oracle.com/en/database/oracle/oracle-database/19/ladbi/about-oracle-linux-and-the-unbreakable-enterprise-kernel.html
  46. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmin/overview-oracle-exadata-deployment-assistant.html
  47. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/monitoring-oracle-vms-oracle-enterprise-manager.html
  48. https://blogs.oracle.com/exadata/post/software-2025-sep
  49. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/verifying-and-modifying-link-speed-client-network-ports-x7-and-later.html
  50. https://www.oracle.com/database/technologies/exadata/software/memoryaccelerator/
  51. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmr/extending-and-multi-rack-cabling-guide-exadata-database-machine-dbmmr.pdf
  52. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/exascale.html
  53. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/exadata-secure-rdma-fabric-isolation.html
  54. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmsq/network-security-considerations.html
  55. https://www.oracle.com/database/technologies/exadata/software/smartscan/
  56. https://www.oracle.com/database/technologies/exadata/software/storageindexes/
  57. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/storage-index.html
  58. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/flash-cache.html
  59. https://www.oracle.com/docs/tech/database/exadata-smart-flash-cache.pdf
  60. https://www.oracle.com/technetwork/server-storage/engineered-systems/exadata/exadata-dbmachine-x3-2-ds-1855384.pdf
  61. https://www.oracle.com/database/technologies/exadata/software/hcc/
  62. https://www.oracle.com/a/ocom/docs/database/hybrid-columnar-compression-brief.pdf
  63. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/exadata-hybrid-columnar-compression.html
  64. https://www.oracle.com/database/technologies/exadata/hardware/storageservers/
  65. https://www.oracle.com/us/solutions/exadata-smart-flash-cache-366203.pdf
  66. https://blogs.oracle.com/maa/more-than-just-redundant-hardware-exadata-maa-and-ha-explained-part-ii-the-exadata-storage-cell
  67. https://www.oracle.com/a/tech/docs/exadata-maa-wp.pdf
  68. https://blogs.oracle.com/maa/more-than-just-redundant-hardware-exadata-maa-and-ha-explained-part-iii-roce-fabric-human-error
  69. https://docs.oracle.com/en/database/oracle/oracle-database/21/haovw/oracle-rac-and-exadata.html
  70. https://docs.oracle.com/cd/E76475_01/html/E76483/golcw.html
  71. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmsq/security-guide-exadata-database-machine-dbmsq.pdf
  72. https://www.oracle.com/at/a/ocom/docs/engineered-systems/exadata/exadata-x11m-ser-ds.pdf
  73. https://docs.oracle.com/en-us/iaas/exadata/doc/ecc-whats-new-in-exadata-cloud-at-customer-gen2.html
  74. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmin/index.html
  75. https://docs.oracle.com/en-us/iaas/exadata/doc/ecc-site-requirements.html
  76. https://www.oracle.com/a/ocom/docs/engineered-systems/exadata/exadata-x11m-ds.pdf
  77. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmsq/using-oracle-ilom-secure-management.html
  78. https://docs.oracle.com/en/database/oracle/oracle-database/19/inmem/intro-to-in-memory-column-store.html
  79. https://www.oracle.com/news/announcement/oracle-helps-customers-achieve-extreme-availability-and-performance-for-mission-critical-and-agentic-ai-applications-2025-08-07/
  80. https://docs.oracle.com/en/solutions/migrate-apps-to-apex/index.html
  81. https://blogs.oracle.com/exadata/post/exadata-technology-series-smart-flash-cache-part-i-a-recap
  82. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/hybrid-columnar-compression.html
  83. https://www.oracle.com/a/ocom/docs/engineered-systems/exadata/dsc-on-x8m-openworld.pdf
  84. https://blogs.oracle.com/database/oracle-database-23ai-vector-search-spatial-graphs-and-grounded-generative-ai-in-enterprise-data-systems
  85. https://www.oracle.com/database/ai-vector-search/
  86. https://blogs.oracle.com/exadata/post/exadata-x11m
  87. https://blogs.oracle.com/exadata/post/exadata-resource-management-consolidation-meets-high-performance
  88. https://blogs.oracle.com/exadata/post/exadata-ai-smart-scan-deep-dive
  89. https://www.oracle.com/us/assets/lifetime-support-technology-069183.pdf
  90. https://www.oracle.com/contracts/docs/hardware-systems-support-policies-069182.pdf
  91. https://support.oracle.com/knowledge/Oracle%20Cloud/3086798_1.html
  92. https://blogs.oracle.com/exadata/post/exadata251
  93. https://blogs.oracle.com/exadata/post/exadata-software-update-tips-and-best-practices
  94. https://www.oracle.com/a/ocom/docs/engineered-systems/exadata/exadata-support-ds.pdf
  95. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/exadata-system-health-checking-exachk.html
  96. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/understanding-rolling-and-non-rolling-updates.html
  97. https://blogs.oracle.com/exadata/post/exadata-patchmgr-making-critical-exadata-updates-effortless
  98. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/sagug/upgrading-oracle-exadata-storage-server-software.html
  99. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/maintenance-guide-exadata-database-machine-dbmmn.pdf
  100. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmmn/exadata-patchmgr-update-utility.html
  101. https://docs.oracle.com/en/engineered-systems/exadata-cloud-service/ecscm/ecs-migrate-to.html
  102. https://docs.oracle.com/en-us/iaas/exadata/doc/ecc-out-of-place-upgrade-gen1-to-gen2.html
  103. https://blogs.oracle.com/observability/right-size-exadata-host-ops-insights
  104. https://blogs.oracle.com/exadata/post/exadata-x8m
  105. https://docs.oracle.com/en-us/iaas/exadatacloud/exacs/ecs-patch-update.html
  106. https://docs.oracle.com/en/engineered-systems/exadata-cloud-at-customer/ecccm/ecc-vw-maint-hist.html
  107. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/25.1-requirements.html
  108. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/25.2-requirements.html
  109. https://docs.oracle.com/en/engineered-systems/exadata-database-machine/dbmso/ai-vector-search-enhancements.html
  110. https://blogs.oracle.com/coretec/post/oracle-database-monthly-news-june-2025-quick-links
Add your contribution
Related Hubs
User Avatar
No comments yet.