Why results stop at 500 rows

DevOpsPad is built for looking at data, not moving it. Each query result is a preview of at most 500 rows or documents and at most 8 MiB of encoded data, and every query has a 20-second timeout. When a result is cut short, the footer tells you why.

To see more, narrow the query or page through it with a range filter, LIMIT/OFFSET, MongoDB batches or Redis cursors.

Limits by engine

Engine	Row or item limit	Size limit	What happens at the limit
MySQL	500 rows	8 MiB	Stops and marks the result partial
ClickHouse	500 rows (max_result_rows=500, overflow mode break)	8 MiB response	Server stops returning rows; max_execution_time=15
MongoDB	limit 1–500 (default 100)	8 MiB	Use Previous batch / Next batch
Redis	Whatever the command returns	8 MiB	Response rejected with advice to use bounded ranges
Kafka	500 partitions	—	Stops and marks the result partial
Kubernetes / K3s	500 objects	8 MiB	Stops and marks the result partial

The grid itself shows 50 rows per page within the fetched result.

Messages you may see

• Result limit reached — the 500-row cap was hit and more rows exist.
• “Preview reached the 8 MiB limit. Narrow the query or select fewer fields.” — the size budget was hit before 500 rows.
• More documents available — MongoDB has another batch. See Page through MongoDB results.
• “Redis response exceeds the 8 MiB preview limit; use GETRANGE, bounded ranges, or SCAN with a smaller COUNT” — the whole response was too large to show.

For MySQL, DevOpsPad reads one row past the limit, so it only reports truncation when another row actually exists.

Why there are limits

The app runs on your laptop and talks to production systems. Fixed limits keep one careless SELECT * from freezing the window or pulling gigabytes over a VPN, and they make the read-only promise easier to keep: you inspect, you do not bulk-extract.

These are preview limits, not hard memory ceilings. A driver may allocate a large row or response before DevOpsPad checks its size.

How to see more

• SQL: add WHERE, select fewer columns, or page with ORDER BY id LIMIT 500 and a WHERE id > last_id range.
• MongoDB: use a range filter on an indexed field for deep pagination; offset batches can shift under concurrent writes.
• Redis: use SCAN with COUNT, HSCAN/SSCAN/ZSCAN, LRANGE or GETRANGE on large values. See Browse Redis keys with SCAN.
• Kubernetes: query one namespace at a time.

Exports have the same limit, because they write what was fetched. See Export query results to CSV.

Timeouts

Every query has a 20-second outer timeout. Driver connections have their own shorter timeouts; Kubernetes API requests use 15 seconds. Cancelling stops the client request, but a server may keep working until its own timeout. See Run and cancel a query.