iblync · Replication
Real time database replication: changes applied as they happen, not overnight
Real time database replication streams each change made in a source database to a target system shortly after it happens, instead of copying whole tables on a schedule. iblync, Interkey's own platform, does this between different kinds of system, with throttling so production keeps its capacity and queueing so a target outage does not lose changes. It is quoted per project and delivered from Riyadh.
In short
- What it is: changes in a source database are streamed to a target continuously, so the target stays close behind the source.
- What iblync adds: throttling and resource control, and queueing while a target is unavailable, with the queue processed on recovery.
- Used for: a reporting or search copy, a standby copy, a bridge during migration, and two live systems kept in step.
- Works between: the relational sources and the NoSQL and other targets listed on iblync's website.
- How it is bought: by quotation, by source system, target system and data volume.
What "real time" should mean for your project
iblync's website describes streaming data changes in real time with minimal latency. It states no fixed delay, and neither will we on this page, because the delay on your system depends on things we can only see in scoping:
- how fast changes arrive at the source at peak;
- how the changes are captured from that source, and what the source allows;
- how much capacity replication may take from production;
- how quickly the target can apply the changes, including its indexes.
Read more
So start from the business question: how far behind may the target be before someone is misled? For a dashboard it may be minutes. For two systems that both take orders it is much tighter. Write that number down, and agree how lag is measured, before choosing anything.
Where real time database replication is used
| Use | What replication does | What to settle first |
|---|---|---|
| Reporting or search copy | Feeds a separate copy, for example a search index, so heavy queries stay off production | How stale the copy may be, and what happens to deletes |
| Bridge during migration | Keeps the new system current while the old one still takes transactions | How long parallel running lasts and what triggers cutover |
| Two live systems | Keeps shared records in step, sometimes in both directions | Which system owns each record, and conflict rules |
| Standby copy | Maintains a second copy of a critical database elsewhere | Recovery expectations; replication is not a backup |
Replication that has no end date is an integration, not a project task. It needs an owner, monitoring and change management.
Flow control: how replication avoids hurting production
The usual failures of replication are not about capturing changes. They happen around the edges:
- The target is unavailable. A design that fails when the target is down loses changes or stops the pipeline. iblync queues changes while the target is away and processes the queue when it recovers.
- Replication competes with production. After an outage, catching up can load the source at the worst moment. iblync provides throttling and control of resource consumption.
- Interrupted work. iblync's site states that an interrupted migration resumes where it left off.
- Nobody watches the lag. If no one measures how far the target is behind, the first sign of trouble is a user reporting missing data. Agree the measure and the alert in scoping.
Our change data capture guide compares the ways changes are captured and what each asks of the source.
Relatedchange data capture guide
Sources and targets listed on iblync's website
| Role | Systems listed |
|---|---|
| Sources (relational) | Including MySQL, PostgreSQL, SQL Server and IBM DB2 |
| Targets | Including Couchbase and Neo4j, and other NoSQL and search systems |
| Also listed | Apache Kafka, Apache Spark, Amazon S3 |
The site says custom connectors are built for systems it does not yet cover. Confirm your exact versions with us before you commit; support is settled per engagement.
Buying real time replication from Interkey in Saudi Arabia
iblync is an Interkey product, delivered with the team that operates it. It is priced by quotation against the source systems, the target systems and the data volume in scope; there is no published price list.
To start, send us the source and target, a rough data volume and how much of it changes each day. Where the platform runs, and whether data leaves your environment, are settled before the first byte moves. If both systems come from one vendor that supplies a supported native path, use that path first; it is usually cheaper.
Questions buyers ask
How fast is real time replication?
iblync's website says it streams changes in real time with minimal latency, and gives no fixed delay for your system. The delay depends on your change rate, the source, the target and the capacity replication may use. We agree a target and how to measure it in scoping.
Does replication slow down our production database?
It can if it is left unlimited. iblync includes throttling and control of resource consumption so replication does not compete with production for capacity.
What happens if the target goes offline?
Changes queue while the target is unavailable and are processed when it recovers, so they are not lost.
Is replication the same as a backup?
No. A replica copies mistakes and deletions as well as good data. Keep backups, and treat a replica as part of availability planning.
Can we replicate from SQL Server or MySQL to a NoSQL system?
Both are listed as sources on iblync's website, and NoSQL and search systems as targets. How far the data models differ decides how much transformation sits between them.
Where can iblync run?
iblync's site says it deploys on any cloud platform or on-premises. We confirm the deployment for your project before work starts.
Next
Related pages
Next step
Scope a replication workload
Tell us the source and target systems, the data volume, how much of it changes each day, and how far behind the target may be. The reply comes from the team that would run it.
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