APromptAura

#performance

Top prompts tagged performance.

0
p/codingAauraNovice4 days agohas example⁂auto-checked

Explain a suspected memory leak from heap snapshots and propose a fix

I suspect a memory leak in service_name (language/runtime). I'll describe symptoms and, if available, heap snapshot diffs or growth-over-time data. Help me: 1. Distinguish a true leak (unbounded growth, objects that should be freed but aren't) from expected memory behavior (cache warming, GC not having run yet, connection pool at steady-state size). Ask what would confirm this if the data I gave isn't enough. 2. From the object types/counts that are growing, infer the likely retention path (what's holding a reference and preventing collection): common culprits are event listeners never removed, closures capturing large scope, unbounded caches/maps, or a connection pool not releasing. 3. Propose the smallest code change to fix the retention, and explain specifically why it breaks the reference chain. 4. Suggest one verification step (a metric to watch, a snapshot to retake) that would confirm the fix worked in production, not just in a local repro. 5. If the leak is architectural (e.g. a fundamentally unbounded cache with no eviction), say so plainly rather than proposing a band-aid. Symptoms/data: symptoms Relevant code: code
ClaudeGPT#debugging#performance#memory-leak
0
0
p/codingAauraNovice4 days agohas example⁂auto-checked

Interpret a flamegraph and turn it into a prioritized optimization list

I'm pasting a description or text export of a CPU flamegraph / profiler output for service_name. Help me turn it into action, not just commentary: 1. Identify the widest frames (highest self-time, not just total time) and name the actual function/operation, not just 'this looks slow'. 2. For each of the top 3 hotspots, give a hypothesis for WHY it's expensive (algorithmic complexity, unnecessary allocation, serialization overhead, lock contention, N+1 pattern) based on the function names and call context I give you. 3. Rank the 3 by expected impact-to-effort ratio, not just raw self-time, factoring in whether the fix is a one-line change or a redesign. 4. For the top-ranked one, propose a concrete fix and estimate (rough order of magnitude, not precise) how much it should reduce that frame's cost. 5. Note any hotspot that's actually expected/acceptable (e.g. a crypto operation that's inherently expensive) so I don't waste time optimizing something that's already near-optimal. Profiler output / description: profile_data Language/runtime: language Context (request volume, latency budget): context
ClaudeGPT#performance#profiling#optimization
0
0
p/codingAauraNovice4 days agohas example⁂auto-checked

Diagnose and rewrite a slow SQL query with execution-plan reasoning

This query is too slow. I'll give you the query, the schema (with indexes), and the EXPLAIN/EXPLAIN ANALYZE output. Do this: 1. Read the execution plan and identify the specific operation that dominates cost (seq scan, nested loop, sort spill, etc.), quoting the relevant line from the plan. 2. Explain in plain terms why the planner is choosing that path given the current indexes and statistics. 3. Propose the smallest change that would fix it, ranked: (a) query rewrite with no schema change, (b) new index, (c) schema change, in that order of preference, and say why you're not jumping straight to an index if a rewrite would do. 4. Show the rewritten query or CREATE INDEX statement. 5. Predict how the plan should change after your fix, and tell me what to check in the new EXPLAIN output to confirm it worked. 6. Flag if the fix could hurt a different query that shares the same table (e.g. a new index that slows down writes on a hot table). Database: {{database, e.g. Postgres 15}} Query: query Schema/indexes: schema EXPLAIN output: explain_output
ClaudeGPT#sql#performance#database
0