What is an upgrade package in managed services?
Upgrading is unavoidable. With fixed packages and a phased programme per DTAP environment, the cost and the schedule of an upgrade are settled in advance.
An upgrade package is a pre-defined package for the version upgrade of a data environment, with a fixed price per component and a fixed programme that surfaces risks step by step before they reach production. Upgrading regularly is not a matter of choice: outdated versions of Qlik Sense (client-managed), erwin Data Modeler or Talend, and of the drivers and connectors that link pipelines to Snowflake, lose support, security patches and compatibility with new integrations. Whoever waits too long upgrades anyway in the end, but under time pressure and with more risk.
Why upgrades are postponed so often
In my experience organisations rarely postpone upgrades out of reluctance. Usually it is uncertainty about the cost and the impact. An upgrade seldom touches a single system: a new Snowflake connector affects the ETL pipelines, a Talend upgrade the Remote Engine or the Runtime, a Qlik upgrade dashboards that have run unchanged for years.
On top of that, teams often wait for the perfect moment. That moment does not come, because there is always a reporting cycle, a peak period or a project that cannot be interrupted. An upgrade that is planned and phased in advance solves that too: the schedule takes the organisation into account from the start, instead of waiting for a quiet period that never arrives.
Fixed packages, fixed prices
At cimt Managed Services upgrades are defined packages with a fixed price per component: the platform itself, the underlying pipelines, the integrations and the dashboards or reports that depend on them. For an integration layer on Talend, for example, that means a separate price for Studio, the Runtime and the Remote Engine, each with its own impact and test cycle. The cost of a full upgrade cycle is clear in advance, instead of an hourly bill that only becomes visible afterwards.
A fixed programme, step by step
Alongside the fixed prices, cimt works with a fixed upgrade programme that makes risks visible as early as possible. After the upgrade, each environment in the DTAP sequence (development, test, acceptance, production) is tested first and then runs for a while, a week for example, before the next environment is upgraded. If a problem appears, it is solved in the environment where it arose, not pushed on to production. This is what such a programme looks like over time:
| Week | Environment | Servers |
|---|---|---|
| 1 | Development | DEV-ESB-01 |
| 2 | Monitoring | no upgrade |
| 3 | Test | TST-ESB-01 TST-DI-01 TST-ESB-02 |
| 4 | Acceptance | ACC-ESB-01 ACC-DI-01 |
| 5 | Production, part 1 | PRD-ESB-01 PRD-DI-01 |
| 6 | Production, part 2 | PRD-ESB-02 PRD-DI-02 |
| 7 | Monitoring | no upgrade |
Each upgrade is scheduled at the start of the week, so the environment runs for the rest of that week before the next phase starts. The week without an upgrade after DEV-ESB-01 is not a delay but extra monitoring time: the first upgraded environment runs longer, and only when nothing comes up is the test environment next. At the end, production runs under extra supervision for one more week.
What this delivers
The goal is not just a successful upgrade, but an upgrade whose cost and schedule are fixed in advance and where a problem in test never becomes a surprise in production. That is the difference between upgrading as last-minute firefighting and upgrading as a fixed part of operations. An upgrade package is separate from an SLA for the data environment or a retainer, but fits well alongside them: the SLA covers day-to-day management, the upgrade package the planned step to a new version.
Would you like to know what an upgrade package means for your data environment? Get in touch and we will map out together which components an upgrade touches.
About the author
Frequently asked
About upgrades
What does an upgrade package cost?
That depends on the components the upgrade touches. The platform, the pipelines, the integrations and the reports each have their own fixed price, so the cost of a full upgrade cycle is clear in advance.
Why not simply upgrade whenever there is time?
That moment rarely arrives on its own. Without a fixed programme there is also no phased check per DTAP environment, so problems only surface in production.
How long does an environment run before the next one is upgraded?
That depends on the situation. About a week is usual, to test the environment and let it run before the next step follows.
What happens if the test environment turns up problems?
Then the next environment in the DTAP sequence only starts once the problem is solved. That is the core of the fixed programme.
Can an upgrade package be taken separately from an SLA or retainer?
Yes. An upgrade package is a defined, one-off project, while an SLA and a retainer are ongoing services. They are often combined.
Further reading
Ready to apply this?
Book a conversation with cimt and see how these insights fit your data foundation.