# Using DataPrime to troubleshoot common query issues

## Common issues and fixes

### 1. “My query is running too slowly”

If your query feels sluggish, consider the following optimizations:

#### 1.1 Prefer exact key comparisons over fuzzy matching

Use direct comparisons like:

```dataprime
filter my.key == 'hello'
```

instead of:

```dataprime
filter my.key.contains('hello')
```

Exact matches are significantly faster because they can take advantage of indexing. Use fuzzy logic only when necessary.

#### 1.2 Narrow your timeframe

Querying large time ranges (weeks or months) increases scan volume and slows results. Reduce your time window when troubleshooting:

- Use the time picker to select only what’s necessary.
- Use relative time queries where applicable.

#### 1.3 Avoid expensive transformations during querying

DataPrime supports dynamic transformations, but these come at a performance cost. Move complex formatting or parsing into **Parsing Rules** during ingestion to reduce query-time processing.

#### 1.4 Filter early

Always reduce your working set before transforming or aggregating. For example:

```dataprime
source logs
| filter name == 'Chris'
| groupby path aggregate count()
```

Filtering first minimizes the data the engine needs to process downstream.

______________________________________________________________________

### 2. “Scan limit exceeded”

You’ve hit the scan cap for your query. This is a guardrail to prevent excessive resource usage.

#### Fixes:

- **Switch to “All Logs Mode” or “All Traces Mode”**: These modes query your full archive (e.g., S3) where scan limits are much higher.
- **Simplify your query**: Remove unnecessary transformations, filters, or nested expressions that force full scans.

For full context, see: [Fair Usage Limits](https://coralogix.com/docs/dataprime/language-reference/limitations/index.md).

______________________________________________________________________

### 3. “The AI Query got my query wrong”

Coralogix AI Query is powered by a large language model and AI can be wrong sometimes. Try rewording your prompt.

#### Fix:

Refine your prompt by:

- **Specifying the exact keypaths**
- **Being concise**
- **Using domain-specific language when helpful**

______________________________________________________________________

### 4. “I get a deprecated warning when I use a command”

DataPrime evolves constantly. Deprecated functions are still supported, but should be updated.

#### Fix:

- Visit the [Functions Reference](https://coralogix.com/docs/dataprime/language-reference/functions-reference/index.md) or [Commands Reference](https://coralogix.com/docs/dataprime/language-reference/commands-reference/index.md).
- Search for the deprecated command and use the suggested replacement.

This ensures your queries stay future-proof.

______________________________________________________________________

## Expected output

You should see:

- Noticeable speed improvements in query results after filtering or narrowing time.
- Errors related to “Scan limit exceeded” disappear in archive modes.
- AI-generated queries that are more accurate with better prompts.
- Warnings resolved after replacing deprecated syntax.

______________________________________________________________________

## Common pitfalls

- Using `contains()` or regex matching as a default—these are slower than direct comparisons.
- Running queries over large time ranges unnecessarily.
- Assuming AI queries will always “just work”—be specific.
- Ignoring deprecation warnings, which can eventually cause query failures.