Vertice

Technical Deep Dive: Adding X10 Storage Cells to an Existing Oracle Exadata Environment

Conor Kehoe — Mon, 06 Apr 2026 09:30:00 +0000

Author: Jason Beattie

Objective

This post details the end-to-end process of expanding an existing Oracle Exadata environment by integrating two additional X10 storage cells. The steps include validating system compatibility, creating grid disks, integrating with Oracle ASM, and verifying post-implementation health and capacity.

Note: Images and specific IP Related Details has been masked for security reasons. Please reach out if you have any questions on this blog post.

System Overview

Environment Setup (Before Expansion):

Existing Storage Cells: Cell01, Cell02, Cell03
Database Nodes: DB01, DB02
ASM Disk Groups: DATA, DONOTUSE, RECOVER
Redundancy Mode: High

Storage Cells Added:

New Cell A
New Cell B

Note: All hostnames and IPs referenced in this blog are anonymized for confidentiality.

1. Initial Validation

Before adding the new cells, we validated the following:

Network Reachability: Ensured new storage cells were reachable from both DB nodes via public and private interfaces.
Image Consistency: Confirmed all storage servers were running matching image versions.
CLI Tools: Verified that CellCLI was operational on both new storage cells.

2. Grid Disk Provisioning

On each of the new cells, grid disks were manually created using CellCLi. Each physical disk was divided among the ASM disk groups according to defined sizing policies.

Sample Commands (Executed via CellCLI):

create griddisk DATA_CD_00_ celldisk=CD_00_, size=8.111328125T create griddisk DONOTUSE_CD_00_ celldisk=CD_00_, size=2.77734375T create griddisk RECOVER_CD_00_ celldisk=CD_00_, size=1.4443359375T

This pattern was repeated across six celldisks per cell for both new storage servers.

3. ASM Integration

With grid disks created, the next step was to integrate them into ASM. First, we validated new disks using:

asmcmd lsdsk --candidate

Then, disks were added to each disk group:

DATA

ALTER DISKGROUP DATA ADD DISK 'o//DATA_CD_00_', ... 'o//DATA_CD_05_';

RECOVER

 ALTER DISKGROUP RECOVER ADD DISK 'o//RECOVER_CD_00_', ... 'o//RECOVER_CD_05_';

DONOTUSE

ALTER DISKGROUP DONOTUSE ADD DISK 'o//DONOTUSE_CD_00_', ... 'o//DONOTUSE_CD_05_';

Each operation triggered an automatic rebalance operation handled internally by ASM.

4. Post-Integration Checks

Once disks were added, we ran several validations:

ASM Usage Verification

SELECT NAME, TOTAL_MB/1024 AS TOTAL_GB, FREE_MB/1024 AS FREE_GB FROM V$ASM_DISKGROUP;

Verified all newly added disks were visible

Image Redacted.

Confirmed that rebalance operations completed successfully.
Checked logs for errors (none found in ASM or alert logs).

5. Capacity Results

Disk Group	Capacity Before (TB)	Capacity After (TB)
DATA	13.59	25.16
DONOTUSE	5.24	8.75
RECOVER	2.72	4.35

Total usable capacity saw a significant increase across all critical disk groups.

6. Summary

Task	Status
Image version check	✅ Completed
Networking configuration	✅ Successful
Grid disk creation	✅ Completed
Disk integration into ASM	✅ Rebalance successful
Capacity validation	✅ Verified

Conclusion

The integration of two new X10 storage cells into the existing Oracle Exadata environment was completed successfully. With zero errors during validation and over 80% storage capacity gain in core disk groups, this expansion lays the foundation for enhanced performance and future scalability.

connect@verticecloud.com

Jason Beattie

Senior Architect and Cloud Operations Manager

The post Technical Deep Dive: Adding X10 Storage Cells to an Existing Oracle Exadata Environment appeared first on Vertice.

Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 2

Conor Kehoe — Mon, 23 Mar 2026 10:30:00 +0000

Author: Philip Godfrey

In the previous blog, we introduced the Oracle Vision and creating a Custom Vision model.

The second blog in the series will work through creating and training a Custom Vision model using our previously labelled dataset.

What is Oracle Vision?

One of many AI Services introduced by Oracle, OCI Vision serves as an advanced AI tool aimed at facilitating extensive image analysis powered by deep learning algorithms.

It streamlines the process of integrating image and text recognition features into applications, eliminating the need for specialised knowledge in machine learning.

Accessing Oracle Vision

In OCI, navigate to Analytics & AI, and under the AI Services subheading, you will find Vision.

Once you click here, providing the necessary pre-requisites have been granted, you should see the overview of Oracle Vision.

The focus of this blog will be Custom models, so we need to navigate to Projects on the left-hand side of the menu:

Custom Models are available in the form of Image Classification and Object Detection.

In the previous blog we created a project, make sure to look back on that if you missed it. Blogs before that explained how we created our labelled dataset, that we will use to train a custom model on.

Note: Minimum Number of Labelled Images for each Label for a Custom Model is 10

There are several steps to creating and training a model including:

Type: Object Detection or Image Classification
Training Data: Existing dataset or Create New Dataset (this will take you through to Data Labelling)
Data Source: Data Labelling service (previously labelled data) or Object Storage
Dataset: Select dataset from desired source

Optional Advanced Settings

Test Dataset: select dataset from object storage bucket
Validation Dataset: select dataset from object storage bucket

Train Model

The next stage in the process of to train a model – for this there are some required fields you need to populate including:

Model Display Name: provide a unique display name for your model, which helps when collaborating with team members

Model Description (optional): optional to provide a description of the model, which helps provide extra context to the model

Training Duration: select from one of three options:

Recommended Mode
Quick Mode
Custom

Which training duration should I choose?

The most appropriate training duration depends on the complexity of your problem, the typical number of objects in an image, the volume of data you are incorporating and budget, along with many other factors.

A summary of each training mode can be found below, to help you determine what is most appropriate

Train Model

As this is a first pass of the model, we have decided to go for Quick mode, with the view we can get an early indication of model performance on a model that is not fully optimized.

Create and Train

The final stage of the process is a review of all steps completed so far.

This provides a summary of the training data asset information, along with Model Information (model name/ type / training duration)

If you need to amend any of these attributes, you can select Edit to update these attributes.

However, if all is as expected, click Create and Train to kick-off the model training process.

The next blog in the series will focus on creating and training a Custom Vision model in OCI.

Once completed, you will see the model has been created and trained successfully.

Interested to know more?

We’d be more than happy to discuss your requirements or questions, please get in touch here.

connect@verticecloud.com

Philip Godfrey

Principal AI & Data Analytics Consultant

The post Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 2 appeared first on Vertice.

Is Your OCI Build Choking Your Own Growth

declanryan — Fri, 13 Mar 2026 15:00:55 +0000

Author: Jason Beattie

Getting support tickets about load balancer limits? The problem probably isn’t Oracle Cloud. It’s the architecture underneath it.

A surprising number of Independent Software Vendors running multi-tenant platforms on Oracle Cloud Infrastructure eventually receive a message that looks something like this:

“Your tenancy has reached its load balancer limit.”

For many organisations, this moment arrives during growth.

A new customer needs onboarding. A new environment needs provisioning.

And suddenly the infrastructure that once felt scalable… isn’t.

At first glance, it’s easy to assume the platform is the problem. Limits exist in every cloud provider, after all.

But in the vast majority of cases, the cloud isn’t the blocker.

The architecture is.

More specifically, it’s often the architecture that was designed early in the journey – sometimes by internal teams, sometimes by external providers – when the platform only had a handful of customers and scale wasn’t yet a concern, or during a rushed migration to the cloud where immediate delivery took priority over long-term scalability.

The Pattern That Quietly Breaks ISV Platforms

One pattern appears again and again in hosted enterprise software environments – particularly platforms built around applications such as Primavera, Oracle E-Business Suite, JD Edwards, PeopleSoft, Unifier, Aconex, and IBM Maximo.

Each customer environment is exposed through its own port.

At first this feels logical. It separates tenants cleanly and allows environments to be provisioned quickly.

It works perfectly when there are five customers.

It works when there are ten.

But by the time platforms reach dozens of customers, the underlying architecture begins to show its age – and organisations start encountering limits that were never meant to be part of the growth journey.

REALITY CHECK
Cloud limits rarely stop platforms from scaling. Architectural patterns that grow infrastructure alongside customers eventually do.

The Question ISVs Should Be Asking

When infrastructure limits begin appearing during growth, the first reaction is often to request higher quotas or deploy additional resources.

But the more important question is this:

Is the architecture designed for where the platform is going – or where it started?

Modern multi-tenant platforms should not require new infrastructure every time a customer is onboarded.

When they do, operational complexity increases alongside the business. Eventually the platform itself becomes the constraint. The good news is that these issues are often solvable with a careful architectural review.

connect@verticecloud.com

Jason Beattie

Senior Architect and Cloud Operations Manager

The post Is Your OCI Build Choking Your Own Growth appeared first on Vertice.

Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 1

declanryan — Mon, 09 Mar 2026 12:36:02 +0000

Author: Philip Godfrey

In the previous blog, we introduced the Oracle Vision service, explaining what it is, any necessary pre-requisites to enable the service, and exploring utilizing pre-trained models within the Vision service.

The next blog in the series will work through creating a Custom Vision model in Oracle Vision.

What is Oracle Vision?

One of many AI Services introduced by Oracle, OCI Vision serves as an advanced AI tool aimed at facilitating extensive image analysis powered by deep learning algorithms.

It streamlines the process of integrating image and text recognition features into applications, eliminating the need for specialised knowledge in machine learning.

Pre-Requisites

To access Vision, there are several steps that need to be taken by an Administrator within your tenancy, including necessary groups, policies and dynamic groups. A full list of requirements can be found here -> About Vision Policies

Accessing Oracle Vision

In OCI, navigate to Analytics & AI, and under the AI Services subheading, you will find Vision.

Once you click here, providing the necessary pre-requisites have been granted, you should see the overview of Oracle Vision.

This has lots of useful resources for you to utilize, including video tutorials introducing the service in simple and easy to consume terms, through to more detailed and in-depth release notes and documentation.

On the left-hand side, you have access to multiple areas:

Overview – the page above
Image Classification – utilize pre-built models to assign classes and confidence scores
Object Detection – utilize pre-built models to identify common objects and confidence scores
Face Detection – utilize pre-built models to identify faces from an image
Text Detection – utilize pre-built models to identify text in documents
Projects – where we can create a custom models (Image classification and object detection) trained on your own labelled dataset.

The focus of this blog will be Custom models, so we need to navigate to Projects on the left-hand side of the menu:

Pre-Requisites

Much like the Vision service, there are some additional setup tasks that need to be performed by an Administrator within the tenancy. These are outlined below.

Custom Models are available in the form of Image Classification and Object Detection.

To get started we select Create project.

We are then presented with a form to specify the details on the project, including Name, Description and any optional Tags.

This will then be available as a custom project from the specified compartment:

The next blog in the series will focus on creating and training a Custom Vision model in OCI.

Interested to know more?

We’d be more than happy to discuss your requirements or questions, please get in touch here.

connect@verticecloud.com

Philip Godfrey

Principal AI & Data Analytics Consultant

The post Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 1 appeared first on Vertice.

Modernise Oracle Forms on OCI

declanryan — Tue, 03 Feb 2026 16:28:30 +0000

Author: Jason Beattie

Oracle Forms and Reports continue to power some of the world’s most mission-critical enterprise systems. Yet for many organisations, aging infrastructure, complex upgrades, and rising operational costs are slowing innovation and increasing risk.

With Oracle Forms 14c (14.1.2) fully supported on Oracle Cloud Infrastructure (OCI), enterprises now have a clear, low-risk path to modernize.

Why OCI Is the Right Home for Oracle Forms

Using the Oracle WebLogic Suite UCM Marketplace Image, organizations can deploy a certified, production-ready Oracle stack that is license-inclusive, removing the need for complex BYOL licensing.

The Marketplace image Currently includes:

Oracle Forms 14.1.2 (14c)
Oracle WebLogic Server 14.1.2 (14c)

For the database tier, organisations can leverage a license-inclusive Oracle Database 19c instance, simplifying procurement and ensuring predictable costs.

This approach eliminates manual installations, reduces configuration errors, and accelerates time-to-value. Allowing IT teams to focus on business outcomes instead of infrastructure management.

The Business Benefits

1. Simplified Licensing

License-inclusive middleware and database options reduce compliance risk and operational overhead.

2. Enterprise-Grade Security

Built-in OCI security services, networking controls, and IAM capabilities protect critical workloads.

3. Oracle Support Alignment

Run on certified, fully supported Oracle versions to minimise risk and extend application lifespan.

4. Future-Ready Architecture

Establish a foundation for gradual modernisation, integration, and cloud-native services on your timeline.

Where Vertice Fits In

Migrating Oracle Forms isn’t just about moving servers! it’s about doing it right.

Vertice helps organisations:

Migrate Oracle Forms and Reports to OCI using certified Marketplace images

Upgrade existing environments to Forms 14c
Optimise WebLogic and database configurations for performance, security, and resilience
Define a roadmap toward future modernisation
Vertice supports the migration as a true partner. We work closely alongside our customers, ensuring they retain control throughout the journey.
Our collaborative approach means nothing happens in the dark. All knowledge is shared, decisions are transparent, and your teams stay empowered every step of the way.

OCI allows enterprises to preserve their Oracle Forms investment while gaining the agility, scalability, and security of the cloud. With the right partner, migration becomes predictable, and low risk.

Vertice makes Oracle Forms cloud migration simple. So, your business can move forward into the modern world!

connect@verticecloud.com

Jason Beattie

Senior Architect and Cloud Operations Manager

The post Modernise Oracle Forms on OCI appeared first on Vertice.

Oracle Vision: Harnessing Pre-Trained Models for Smarter Insights

declanryan — Wed, 28 Jan 2026 12:30:15 +0000

Author: Philip Godfrey

One of many AI Services introduced by Oracle, OCI Vision serves as an advanced AI tool aimed at facilitating extensive image analysis powered by deep learning algorithms.

It streamlines the process of integrating image and text recognition features into applications, eliminating the need for specialised knowledge in machine learning.

Pre-Requisites

Accessing Oracle Vision

In OCI, navigate to Analytics & AI, and under the AI Services subheading, you will find Vision.

Once you click here, providing the necessary pre-requisites have been granted, you should see the overview of Oracle Vision.

On the left-hand side, you have access to multiple areas:

Overview – the page above
Image Classification – utilize pre-built models to assign classes and confidence scores
Object Detection – utilize pre-built models to identify common objects and confidence scores
Face Detection – utilize pre-built models to identify faces from an image
Text Detection – utilize pre-built models to identify text in documents
Projects – where we can create a custom models (Image classification and object detection) trained on your own labelled dataset.

Pre-Trained Models

Object Detection

As an example, we can pass in an image from a previous blog series (car parks / spaces) to see how the pre-trained model performs in identifying cars.

If we select Object Detection we can then navigate to a Local File and pass in a png/jpeg file, the pre-trained model is then invoked and objects identified are returned. It has successfully (and confidently) identified all cars in the image with a good level of confidence!

Image Classification

As a further image, we can also call the pre-trained Image Classification model, if we provide another example image, we can see what it has identified:

Exploring the results we can see that Cars have been identified with the greatest confidence (which you would expect) however it has also identified there are Vehicles, Land vehicle, and it has also identified there is a parking lot in the picture! So, we can get lots of useful information from the pre-trained models.

Join our next blog in the series where we cover Custom Models in detail, and how we can create Vision models based on our own data.

Custom Models

There are several pretrained models available for Image Classification, Object Detection and Text Recognition.

Custom Models are available in the form of Image Classification and Object Detection.

Pretrained Models:

Image Classification
Object Detection
Text Recognition

Custom Models:

Image Classification
Object Detection

The next blog in the series will explore creating a Custom model in Oracle Vision and show you how easy it is to get up and running.

connect@verticecloud.com

Philip Godfrey

Principal AI & Data Analytics Consultant

The post Oracle Vision: Harnessing Pre-Trained Models for Smarter Insights appeared first on Vertice.

Adding X10 Storage Cells to an Existing Oracle Exadata Environment

colmmccarthy — Fri, 02 Jan 2026 11:36:45 +0000

Author: Jason Beattie

Objective

System Overview

Environment Setup (Before Expansion):

Existing Storage Cells: Cell01, Cell02, Cell03
Database Nodes: DB01, DB02
ASM Disk Groups: DATA, DONOTUSE, RECOVER
Redundancy Mode: High

Storage Cells Added:

New Cell A
New Cell B

Note: All hostnames and IPs referenced in this blog are anonymized for confidentiality.

1. Initial Validation

Before adding the new cells, we validated the following:

Network Reachability: Ensured new storage cells were reachable from both DB nodes via public and private interfaces.
Image Consistency: Confirmed all storage servers were running matching image versions.
CLI Tools: Verified that CellCLI was operational on both new storage cells.

2. Grid Disk Provisioning

On each of the new cells, grid disks were manually created using CellCLi. Each physical disk was divided among the ASM disk groups according to defined sizing policies.

Sample Commands (Executed via CellCLI):

				
					create griddisk DATA_CD_00_<cellname> celldisk=CD_00_<cellname>, size=8.111328125T create griddisk 
DONOTUSE_CD_00_<cellname> celldisk=CD_00_<cellname>, size=2.77734375T create griddisk 
RECOVER_CD_00_<cellname> celldisk=CD_00_<cellname>, size=1.4443359375T

This pattern was repeated across six celldisks per cell for both new storage servers.

3. ASM Integration

With grid disks created, the next step was to integrate them into ASM. First, we validated new disks using:

				
					asmcmd lsdsk --candidate

Then, disks were added to each disk group:

DATA

				
					ALTER DISKGROUP DATA ADD DISK 'o/<private_ip_set>/DATA_CD_00_<cellA>', ... 'o/<private_ip_set>/DATA_CD_05_<cellB>';

RECOVER

				
					 ALTER DISKGROUP RECOVER ADD DISK 'o/<private_ip_set>/RECOVER_CD_00_<cellA>', ... 'o/<private_ip_set>/RECOVER_CD_05_<cellB>';

DO NOT USE

				
					ALTER DISKGROUP DONOTUSE ADD DISK 'o/<private_ip_set>/DONOTUSE_CD_00_<cellA>', ... 'o/<private_ip_set>/DONOTUSE_CD_05_<cellB>';

Each operation triggered an automatic rebalance operation handled internally by ASM.

4. Post-Integration Checks

Once disks were added, we ran several validations:

ASM Usage Verification

				
					SELECT NAME, TOTAL_MB/1024 AS TOTAL_GB, FREE_MB/1024 AS FREE_GB FROM V$ASM_DISKGROUP;

Verified all newly added disks were visible.

Confirmed that rebalance operations completed successfully.
Checked logs for errors (none found in ASM or alert logs).

5. Capacity Results

Total usable capacity saw a significant increase across all critical disk groups.

6. Summary

Conclusion

connect@verticecloud.com

Jason Beattie

Senior Architect and Cloud Operations Manager

The post Adding X10 Storage Cells to an Existing Oracle Exadata Environment appeared first on Vertice.

AI-Driven Language Narrative in Oracle Analytics

colmmccarthy — Tue, 16 Dec 2025 10:46:50 +0000

Author: Philip Godfrey

Oracle Analytics Cloud is a powerful tool allowing users, both individuals and businesses alike, to explore and understand their data by utilizing advanced and augmented analytic capabilities.

These augmentations, such as Auto-Insights, Explain and AI Assistant integration, utilises Machine Learning and AI elements to provide even more options to assist with Data Storytelling.

One of my favourite features in this space is Language Narrative, which is the focus of this blog, as it’s had some recent and very useful enhancements made available in the Oracle Analytics Cloud July 2025 update.

What is Language Narrative?

Language Narrative in a nutshell provides an AI generated summary of attributes and measures passed to it. It takes your visualisation and generates a summary on-the-fly.

Under the hood, it utilizes Natural Language Processing, a subfield of Artificial Intelligence and Machine Learning, to generate this summary based on the attributes passed to it.

From a user perspective however, like most other features in OAC, its very user friendly, with a simple drag-and-drop interface, and a lot can be achieved in a few clicks.

Using Language Narrative

You can utilise Language Narrative from the grammar pane, using this icon here:

Key Components of Language Narrative

Level of Detail: sliding scale between 1 and 7

o 1: Short summary of visualization

o 4: More detailed, but easily consumed

o 7: Lowest level of detail

Analysis Selection: user can select between two options for analysis:

o Trend

o Breakdown

Language Selection: user can select which language to report:

o English

o French

What’s New?

The July 2025 release of Oracle Analytics provided a key update for authors enabling improved Data Storytelling, opportunities, with the inclusion of three distinct tones:

Factual
Business
Casual

These tones allow users to select the most appropriate tone, based on their audience, which is critical for impactful storytelling, ensuring the insights you have derived can be easily consumed.

Factual – clear and precise

This example uses clear language, quite direct with no element of confusion: “the data shows Review_ID for the following Year_Month values”

If we focus on that first sentence of generated text, and compare it to the other tones:

Business – professional and strategic

Quite like factual, but aligned with more strategic language.

The same example as above now returns. “An analysis of the data reveals Review_ID corresponding to the following Year_Month values”

Casual – conversational and engaging

The same text as above no presents as much more easy-going, maybe like you would use if you were showing an analysis to a co-worker:

“Here’s what the numbers tell us: we’ve got Review_ID for these specific Year_Month values”

It feels very friendly, much more relaxed, and a great option depending on your audience.

When to use each tone?

This decision comes down to you as an individual, but I’d like to share some examples of when I might use it in case it’s helpful depending on your audience.

Interested to know more?

We’d be more than happy to discuss your requirements or questions, please get in touch here.

connect@verticecloud.com

Philip Godfrey

Principal AI & Data Analytics Consultant

The post AI-Driven Language Narrative in Oracle Analytics appeared first on Vertice.

False Positive Financial Services Demo

colmmccarthy — Mon, 01 Dec 2025 11:31:27 +0000

The post False Positive Financial Services Demo appeared first on Vertice.

Predictive Maintenance on Infrastructure Demo

colmmccarthy — Mon, 01 Dec 2025 10:44:30 +0000

The post Predictive Maintenance on Infrastructure Demo appeared first on Vertice.

Vertice

Technical Deep Dive: Adding X10 Storage Cells to an Existing Oracle Exadata Environment

Author: Jason Beattie

Objective

System Overview

1. Initial Validation

2. Grid Disk Provisioning

3. ASM Integration

DATA

RECOVER

DONOTUSE

4. Post-Integration Checks

ASM Usage Verification

5. Capacity Results

Disk Group

Capacity Before (TB)

Capacity After (TB)

6. Summary

Task

Status

Conclusion

Jason Beattie

Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 2

Author: Philip Godfrey

What is Oracle Vision?

Accessing Oracle Vision

Train Model

Which training duration should I choose?

Train Model

Create and Train

Interested to know more?

Philip Godfrey

Is Your OCI Build Choking Your Own Growth

Author: Jason Beattie

“Your tenancy has reached its load balancer limit.”

The architecture is.

The Pattern That Quietly Breaks ISV Platforms

The Question ISVs Should Be Asking

Jason Beattie

Oracle Vision AI: A Beginner’s Guide to Custom Models – Part 1

Author: Philip Godfrey

What is Oracle Vision?

Pre-Requisites

Accessing Oracle Vision

Pre-Requisites

Interested to know more?

Philip Godfrey

Modernise Oracle Forms on OCI

Author: Jason Beattie

Why OCI Is the Right Home for Oracle Forms

The Business Benefits

1. Simplified Licensing

2. Enterprise-Grade Security

3. Oracle Support Alignment

4. Future-Ready Architecture

Where Vertice Fits In

Vertice helps organisations:

Jason Beattie

Oracle Vision: Harnessing Pre-Trained Models for Smarter Insights

Author: Philip Godfrey

Pre-Requisites

Accessing Oracle Vision

Pre-Trained Models

Image Classification

Philip Godfrey

Adding X10 Storage Cells to an Existing Oracle Exadata Environment

Author: Jason Beattie

Objective

System Overview

1. Initial Validation

2. Grid Disk Provisioning

3. ASM Integration

DATA

RECOVER

DO NOT USE

4. Post-Integration Checks

ASM Usage Verification

5. Capacity Results

6. Summary

Conclusion