Oracle’s recently announced Exadata Database Service on Exascale (ExaDB-XS) aims to improve performance for database workloads and reduce costs.
The architecture in Exascale, which is Oracle’s multi-tenant architecture that intelligently allocates pools of compute optimised for its databases, is built on the same appliances – X8M, X9M and X10M – Oracle uses on its Exadata Database Service on Dedicated Infrastructure. Exadata dates back a decade and is Oracle’s infrastructure for its relational database.
What’s new is a focus on artificial intelligence (AI) workloads and vector databases, as well as a pay-to-use cost structure.
Since its X8M appliances became available, Oracle has connected servers to storage via RoCE networking, and Oracle’s storage servers are equipped with its XRMEM memory, accessed via RDMA from NVMe drives and HDDs to form three tiers of storage: hot, warm and cold.
That logic has been adapted to a shared architecture, principally by modifying software management of the databases.
“Before this, every tenant had its own dedicated Exadata compute and storage,” said Kodi Umamageswaran, senior vice-president for Exadata and scale-out technologies at Oracle.
“And we had software called Automatic Storage Management (ASM) that was used to distribute storage between databases,” he said. “Now, a common pool of Exascale compute and storage will take charge of thousands of tenants and millions of databases.”
To sum up, a specific Exascale control plane is used to manage VMs and storage.
Traditionally, Oracle represented each hard drive in the OS via a logical unit number (LUN) and managed all logical assets in the form of files, as data in a database, snapshots, redo logs, clones and storage blocks.
But with Oracle ASM, each file type is shared per each group of drives associated with the database, its recovery files, snapshots and extended storage configuration. That serves to ensure redundancy of the data and its separation between discrete storage volumes.
From now, with Exascale, it will be the file type that determines redundancy. Templates allow for configuration of levels of redundancy and the control plane ensures data isn’t stored on the same drive.
Also, files are divided into 8MB “extents”, as Oracle calls them, a size chosen to optimise performance in the multitenant architecture. “These 8MB fragments are sufficiently large to get good sequential performance when we scan contiguous data,” said Umamageswaran. “And sufficiently small to allow distribution of a database across the storage cloud in order to share the I/O load.”
Extents are allocated to storage buckets via a hash code, and a mapping table indicates which disk they are stored on. “All the extents that reside in a bucket are stored in a redundant manner on three disks on three storage servers to protect against storage outages and service interruptions,” said Umamageswaran.
The maximum number of buckets is fixed at 100,000. “It’s sufficient to share the data between thousands of storage servers and small enough for the correspondence table that’s held in cache in client database servers,” he said.
Performance equal to dedicated infrastructure
This architecture, which uses a RoCE network and RDMA connection protocol, allows for performance similar to dedicated Exadata services. That means throughput of 2,880GBps with generic benchmark workloads and latency of 17µs, with X10M servers.
Such performance derives from on-the-fly data tiering and automates parallelisation of SQL processing, among other things.
Developers can profit from being able to create database clones, entire or “thin”, from a database in production or a snapshot. Redirect-on-write technology reduces storage capacity required for clones because clones share blocks from the original database without saving new data. Most of these functions already exist, however.
In place of groups of disks, admins now see “vaults” attached to VM clusters. These vaults are logical storage containers that pull resources from physical servers attached to VM clusters.
Vaults represent a new, more visible way of working for admins because they are used directly by the Oracle database in place of ASM disk groups.
During deployment, it’s possible to configure VMs, attached vaults, and to provision the database container image in the same way as the virtual cloud network, and its client and backup networks.
Admins can deploy between two and 10 VMs, with each allocated an elastic number of cores per hours in Oracle’s ECPU schema. One vault can be between 300GB and 100TB of storage space.
Less costly, at a smaller scale
Owing to the multitenant architecture, ExaDB-XS will be 95% less expensive than Exadata Cloud Infrastructure X9M Dedicated.
For a base configuration (including licence) that uses a quarter rack X9M – i.e. two database servers for three storage servers – dedicated Exadata Cloud Infrastructure costs €10,000 a month approximately compared with about €330 per month with Exadata Exascale infrastructure.
You have to add running Exadata Database services, which Oracle invoices as processing via RDMA. A base instance of ExaDB-XS costs about €4,000 a month compared with €13,700 for an X9M dedicated instance in the cloud, with 4 OCPU (8vCPU) associated with Exadata Database Service. That’s a saving of about 70%, according to the supplier.
While the quarter rack X9M instance has more storage capacity (190TB), 4 OCPU is not enough to process this amount of data. So, Exadata Exascale may not be less expensive for the same storage capacity.
“The small cost of Exascale allows large enterprises and SMEs to use it for small workloads,” said Umamageswaran.
Analysts agree that Oracle is continuing to innovate in the face of competition from competitors. Key among these are PostgreSQL and its variants, supported by large deployments among the cloud giants, plus MongoDB. Oracle was slow to offer usage-based licensing, but seems intent on catching up.
In future, Exadata Exascale will be the underlying architecture for all database services in Oracle Cloud Infrastructure, according to Umamageswaran.
#Oracle #Exadata #Exascale #Big #architecture #shift #small #workloads
