Solution Overview — Summary
Most database performance problems share two unkind properties: they are visible to users long before they are visible to operators, and the available evidence reaches the operator’s desk far too late and far too coarse. iWatch closes that gap.
By capturing every SQL transaction at wire level and writing it to a searchable repository, iWatch turns “the database was slow at around 10:42 this morning” into a precise, evidenced trail — statement, session, response time, and rows affected — that the team can act on within minutes rather than days.
The business case
| Outcome | How iWatch delivers it |
|---|---|
| Faster incident resolution | Operators see exactly which statement was slow, in which session, at which moment — with timing. Less guessing, fewer escalations, shorter outages. |
| No production risk | iWatch is read-only and passive. It does not modify traffic, does not run inside the database, and does not require any application change. It can only observe. |
| Low total overhead | 1–2% of one CPU core in the typical install. Effectively zero on the database host in the separate-host install. |
| One tool, every database | Informix, Db2, SQL Server, PostgreSQL, MongoDB, SAP ASE, SAP IQ, OneDB. One operational grammar across the estate. |
| Repository scales | Purpose-built for tens of millions of measured transactions per day, retained at decreasing detail over time and queryable in SQL throughout. |
What it removes from the risk register
- “We can’t tell whether the slowness is the database, the network, or the application.”
- “Our investigations take days because we don’t have visibility at this layer.”
- “Adding tracing to production is too risky.”
- “We’re paying for senior consulting time on problems we should be able to triage in-house.”
Why Oninit
Oninit has supported the global Informix community for two decades as “the Down System Specialists.” iWatch is delivered and supported by the same team that gets called when other people’s database systems are down — the engineers who already do this kind of investigation for a living, productised so the customer’s team can do it themselves.
Time to value
| Step | Typical effort |
|---|---|
| Receive distribution, install on a host | Under an hour for a single database. |
| Point at the target instance, capture the first session | Minutes. No application restart, no database restart, no configuration change inside the database. |
| First operator-driven incident triage | Hours. The repository is SQL-searchable; an experienced DBA learns the schema in one sitting. |
| Standing observability across the estate | Days to weeks, depending on estate size and the variety of database engines in scope. |
Next step
A short scoping call to understand the environment and the incidents iWatch would shorten. A working trial against a non-production instance can be running the same week. Call +1 913 732 8892 (US) or +44 20 8133 7529 (UK), or email support.info@oninit.com.