iblync · Heterogeneous migration

Heterogeneous database migration: moving data between different database engines

Move data from one database engine to another, in one pass or in phases, with source and target kept in sync until you decide to switch.

  • Built and supported by Interkey An Interkey product, delivered and supported by Interkey
  • One platform for migration and continuous sync The move and the replication that follows it
  • Written quote after one meeting The price is discussed in a meeting about your systems
iblync screen mapping source tables and columns to the target
Table and column mapping in iblync, from the recorded product demo.

A heterogeneous database migration moves data between two different database engines, for example from Oracle to PostgreSQL or from SQL Server to MySQL. The hard part is not the copy: it is mapping types and structures and keeping both systems in step while applications move. iblync, built by Interkey, does the mapping, the load and the continuous sync in one platform.

What changes between two database engines

What changes between two database engines: why each item differs, and what is decided before the move
ItemWhy it differsDecided before the move
Data typesEach engine names and sizes its types in its own way, so a number, a date or a long text in one may have no exact match in the otherThe target type of every column, and how precision and length are kept
Keys and sequencesEngines create new key values in different ways, through sequences or self-numbering columnsHow keys are created on the target, and where numbering continues after the switch
Text encodingA character set that holds Arabic on the source may not match the default on the targetThe target encoding and sort rules, checked with real Arabic text
Dates and time zonesEngines store dates with or without a time zone, and with different precisionWhich time zone each date column holds, and the precision to keep
Empty valuesNot every engine treats an empty string and a NULL value as the same thingOne rule for empty values, applied to every table
Code stored in the databaseProcedures, functions and triggers are written in each engine's own languageWhich code is rewritten for the target: no tool carries it across

None of this belongs to one product: it is what any move between two engines has to settle, whichever tool carries the data. Our Oracle to PostgreSQL migration guide works through these items for one common pair.

What iblync does in a heterogeneous migration

01

Maps tables and columns

Source tables and columns are mapped to the target in low-code workflows, configured by the Interkey team, instead of scripts written for each project. No development team needed on your side.

02

One pass or in phases

Move everything in one migration, or system by system and table by table, in the order your applications move.

03

Continuous sync until the switch

After the initial load, changes on the source keep reaching the target, so both stay in step while applications move. Synchronisation in both directions is supported where both sides accept writes.

04

Controlled load, queued changes

Transfer speed and resource use can be limited, so production is not slowed. If the target stops, changes wait in a queue and are applied when it returns.

Sources and targets

Systems iblync connects as a source or a target, grouped by kind
Kind of systemSystems iblync connects
Relational databasesOracle, SQL Server, PostgreSQL, MySQL
Relational databases on IBM platformsIBM Db2 on LUW, z/OS and IBM i
Document databaseCouchbase
Graph databaseNeo4j
Processing and object storageApache Spark, Amazon S3

Every source and target pair is confirmed in the assessment before the quote. If one of your systems is not on the list, a custom connector is built when needed, as part of the project. The full list is on the iblync connectors page.

How a heterogeneous migration runs

  1. Assess the source and the target.

    Versions, data volume, rate of change and stored code are reviewed, and the pair is confirmed. From you: access to both systems, and someone who knows the applications.

  2. Map tables and types.

    Every table and column gets its target type and the rules from the table above. From you: a decision on each item the mapping raises.

  3. Initial load.

    The existing data is copied to the target, with the load on the source limited. From you: the hours when the source can take the extra load.

  4. Continuous sync.

    Changes on the source keep reaching the target while applications still use the source. From you: the order in which applications will move.

  5. Compare the data.

    Both sides are checked against acceptance checks agreed in advance. From you: the checks that prove the result to your team.

  6. Switch.

    Applications move to the target at the time you choose. From you: the go-ahead, and the downtime window you agreed.

Checking the data and choosing the moment to switch

The last two steps decide whether the migration is accepted, so they are planned at the start, not at the end. Before the first load, your team and ours agree which checks prove the result: row counts per table, totals on the key numeric columns, and samples that include Arabic text, dates and empty values, the items that differ most between engines. iblync's weekly and monthly replication reports give a detailed account of the transfer and show that it is complete, so your teams can see that the old system and the new one match.

Read more

While continuous sync runs, the source stays in service, so the switch can wait until the checks pass and your applications are ready. Many teams run both systems side by side for a period first: our page on the parallel run and cutover plan covers how to set the conditions for ending it, and zero downtime migration covers keeping the service up during the move. The work is set up on iblync and delivered with the Interkey team that operates it.

Watch a move between two engines

A short recording from the iblync product channel: data moving from SQL Server to Couchbase.

A frame from the iblync video: migrating data from SQL Server to CouchbaseData migration from SQL Server to CouchbaseVideo on YouTube, 3:19

What we need from you to quote

Six details. What you do not know yet, we work out together in the meeting.

  • The source engine and its version.
  • The target.
  • The approximate data volume.
  • The daily rate of change.
  • The downtime window you can accept.
  • Where the platform must run.

What iblync does not do

Data, not code

iblync moves data and keeps it in sync. It does not rewrite stored procedures, functions, triggers or application code for the new engine. That is engineering work of its own: we review the code in the assessment and estimate its rewrite separately, so you see it before the migration starts, not after the first load.

Questions about heterogeneous migration

What is heterogeneous database migration?

It is moving data between two different database engines, not upgrading within the same engine. Because the engines differ, each data type, key, encoding and date has to be mapped to its counterpart on the target, and the two systems kept in step while applications move. A plain copy is rarely enough: the mapping decisions are most of the work.

Which database pairs does iblync support?

iblync connects the systems listed on the connectors page, as a source or a target, and your exact pair and versions are confirmed in the assessment before the quote. If one of your systems is not on the list, a custom connector is built when needed, as part of the project.

Can applications keep running during the migration?

Yes, during the continuous sync stage: the source stays in service and its changes keep reaching the target. Applications move when the checks pass and you choose the time, within the downtime window you agreed. Our page on zero downtime migration explains how the switch is planned.

How long does a heterogeneous migration take?

It depends on the data volume, the number of tables, the daily rate of change and how much code is stored in the database. We do not give a general figure, because it would describe another project. The assessment of your source and target gives you a timeline written for your own systems.

Does iblync convert stored procedures?

No. iblync moves data and keeps it in sync; rewriting stored procedures, functions, triggers and application code for the new engine is separate engineering work. We review that code in the assessment, so its rewrite is estimated on its own and you see it before the migration starts.

How is the migration priced?

The price is discussed in a meeting and followed by a written quote, which you can book from the iblync pricing page or the form on this page.

Next step

Plan your move between two engines

Tell us the source engine and its version, the target, the rough data volume and where the platform must run. The price is discussed in a meeting, followed by a written quote.

Required field

By sending this form, you agree to our Privacy Notice.

What happens next

  1. We read your request
  2. An engineer contacts you
  3. You get an estimate or a demo
  • Built and supported by Interkey

    An Interkey product, delivered and supported by Interkey

  • One platform for migration and continuous sync

    The move and the replication that follows it

  • Written quote after one meeting

    The price is discussed in a meeting about your systems

Or directly

+966-11-2180999 [email protected]

Tawuniya Towers, North Tower, 7th Floor, King Fahad Highway, Olaya, P.O. Box 56835, Riyadh 11564, Saudi Arabia

or See supported connectors

Published by Interkey. Last updated . Interkey is registered in Riyadh, Saudi Arabia under commercial registration 1010156897.