Search
30 results for “performance”
Search results
Web Performance & Core Web Vitals
What LCP, INP, and CLS each actually measure, why "good" is defined at the 75th percentile of real page loads rather than the average, and why a fast average can still hide a genuinely bad user experience.
Browser Rendering Pipeline
Why animating transform and opacity can skip layout and paint entirely while animating width or top can't, and how alternating writes and reads to layout properties in a loop forces the browser to recalculate layout over and over.
PostgreSQL Indexes
How B-Tree, Hash, GIN, BRIN, and covering indexes work in PostgreSQL, when to create them, how to inspect usage, and the write-overhead trade-offs at scale.
Redis Caching Patterns
Cache-aside, write-through, and write-behind explained, plus the TTL, stampede, and invalidation pitfalls that show up once Redis caching hits real traffic.
What are the five stages of the browser rendering pipeline, and which ones does a change to a property like width actually have to go through?
The pipeline runs JavaScript, Style calculation, Layout, Paint, and Composite, in that order. Changing a property that affects geometry, `width`, `height`, `position`, forces the browser through every stage: layout has to be recalculated (since the element's size or position changed), then paint (since pixels changed), then composite. That full path is why layout-affecting properties are the most expensive to animate, every frame re-runs the whole pipeline, not just a cheap final step.
Why can animating transform or opacity skip layout and paint entirely, while animating top or width cannot?
transform and opacity don't change an element's geometry, its size or position in the document flow, or repaint its actual pixel content, they only change how an already-painted layer is displayed (moved, scaled, faded) during compositing. Because nothing about the element's layout or pixels actually changed, the browser can skip straight to the composite stage, the cheapest, fastest path in the pipeline, and handle it on the compositor thread. `top` and `width` do change geometry, so there's no way to skip layout, the browser has no shortcut for "the size changed but skip figuring out the new size."
What is layout thrashing, and why does writing then reading a layout property in a loop specifically cause it?
Layout thrashing happens when code alternates writing a style (which invalidates the current layout) and reading a layout-dependent property like `offsetWidth` (which forces the browser to immediately recalculate layout synchronously to answer that read accurately), repeated across many elements in a loop. Each read-after-write pair forces a fresh, synchronous layout recalculation instead of letting the browser batch and defer that work to its normal rendering schedule, because the code demanded an up-to-date value mid-loop. The fix is mechanical: batch every write first, then batch every read afterward, so layout is only recalculated once instead of once per element.
How does PostgreSQL decide whether to use an index?
The query planner estimates the cost of each available access path using table statistics (row count, column cardinality, correlation) gathered by ANALYZE. It compares sequential-scan cost against index-scan cost for the specific query. For small tables, or queries that touch a large fraction of rows, a sequential scan often wins even when an index exists, an index is not free to use if it does not narrow the result much. Run EXPLAIN ANALYZE to see the chosen plan, and re-run ANALYZE after large data changes so statistics stay accurate.
What is the difference between a composite index and two separate single-column indexes?
A composite (multi-column) index on (a, b) stores rows sorted by a, then by b within each a. It efficiently serves queries filtering on a alone, or on a and b together, but not on b alone. Two separate single-column indexes let Postgres combine them via a bitmap AND/OR, which works but is usually slower than one well-ordered composite index for the common query pattern. Column order in a composite index should match the most selective, most-frequently-filtered column first.
When would you choose a partial index over a full index?
When queries only ever filter on a subset of rows, for example WHERE is_active = true, or WHERE deleted_at IS NULL. A partial index (CREATE INDEX ... WHERE condition) only stores entries for matching rows, so it is smaller, faster to scan, and cheaper to maintain on writes than indexing the whole table, at the cost of only being usable when the query's WHERE clause matches (or implies) the index condition.
What is cache stampede and how do you prevent it?
A cache stampede happens when a popular cache key expires and many concurrent requests all miss at once, each falling through to hit the (often slow) origin, database or upstream API, simultaneously, sometimes overwhelming it. Common mitigations: a short-lived lock so only one request repopulates the cache while others wait or serve stale data (the "thundering herd" lock pattern), staggered/jittered TTLs so keys do not all expire at the same instant, and serving stale-while-revalidate, returning the expired value immediately while refreshing it in the background.
What is the difference between cache-aside and write-through caching?
Cache-aside (lazy loading): the application checks the cache first; on a miss it reads from the database, then writes the result into the cache. Writes go to the database only, so the cache can go stale until the next miss or an explicit invalidation. Write-through: every write goes to the cache and the database together (usually the cache write happens synchronously as part of the write path), so the cache is always in sync with the database, at the cost of extra write latency on every mutation.
Why can naive cache invalidation cause a race condition?
If a write invalidates (deletes) a cache key and then updates the database, a concurrent read between those two steps can repopulate the cache with the old value right after it was deleted, leaving stale data cached until the next TTL expiry or invalidation. Ordering matters: update the database first, then invalidate the cache, and even then, a very unlucky interleaving with a concurrent cache-aside read can still occur, which is why short TTLs are used as a safety net rather than relying on invalidation alone for strict consistency.
What do LCP, INP, and CLS each measure, and why are all three needed instead of one overall "speed" number?
LCP (Largest Contentful Paint) measures loading performance, specifically how quickly the largest visible element renders, "good" is 2.5 seconds or less. INP (Interaction to Next Paint) measures interactivity/responsiveness, the delay between a user interaction and the browser's next visual response, "good" is 200 milliseconds or less. CLS (Cumulative Layout Shift) measures visual stability, how much content unexpectedly shifts around during the page's lifecycle, "good" is a score of 0.1 or less. A single overall number couldn't distinguish a page that loads fast but jumps around from one that's stable but slow to interact with, three separate metrics are needed because loading, interactivity, and stability are genuinely different failure modes.
Why does Google measure Core Web Vitals at the 75th percentile of real page loads instead of using the average?
An average can be pulled down by a large number of fast loads on good connections and powerful devices while completely hiding a meaningful tail of slow, frustrating experiences on weaker devices or networks, real users don't experience "the average," they experience their own specific load. Measuring at the 75th percentile means a page only passes if at least three out of four real page loads actually meet the threshold, which is a much more honest bar for "most users get a genuinely good experience" than an average that a handful of very fast loads could distort.
A page has a fast average LCP but users still frequently report the page feeling slow. What could the percentile-based measurement reveal that an average wouldn't?
If a substantial slice of real page loads, say the slowest 25%, badly miss the 2.5 second LCP threshold (a slow connection, a busy device, a cold cache), the average can still look fine because it's dominated by the faster majority, while a real, sizable group of users are having a genuinely bad experience the average is actively hiding. Checking the 75th-percentile value directly (rather than the mean) surfaces that gap: if the 75th percentile is well above 2.5 seconds even though the average looks fine, that's a concrete signal that meaningful numbers of real users are missing the threshold, not a false alarm.
Redis vs Memcached
How Redis and Memcached differ in data structures, persistence, clustering, and threading, and which cache fits which workload.
Database Partitioning
How range, list, and hash partitioning each split one logical table into physical pieces, and why partition pruning depends entirely on the WHERE clause matching partition bounds directly, not on any index.
What is the difference between range, list, and hash partitioning, and when would you choose each?
Range partitioning divides rows by a value falling within a bounded, non-overlapping range (inclusive lower bound, exclusive upper bound), the natural fit for time-series data like logs partitioned by month. List partitioning explicitly assigns specific key values to specific partitions, a good fit when data naturally groups into a known, finite set of categories, like a specific list of counties or regions. Hash partitioning distributes rows by the hash of the partition key modulo a chosen number of partitions, useful specifically when there is no natural range or category to split on and you just need to spread rows roughly evenly across a fixed number of partitions.
A query filters WHERE logdate >= 2008-01-01 against a table range-partitioned by logdate across dozens of monthly partitions. What does partition pruning actually do, and what does it depend on?
Partition pruning lets the planner prove, from the query's WHERE clause and each partition's declared bounds, that some partitions cannot possibly contain a matching row, and it excludes them from the plan entirely rather than scanning and filtering every partition. In this example, `logdate >= 2008-01-01` has no upper bound, so it prunes only the partitions entirely before 2008-01, decades of older monthly partitions are eliminated before execution, while every partition from 2008-01 onward, including all of them up to the present, is still considered and scanned. Pruning down to a single partition would need a bounded predicate on both ends, for example `logdate >= 2008-01-01 AND logdate < 2008-02-01`. This depends entirely on the partition bounds themselves, not on any index, a partitioned table with no indexes at all still benefits from pruning, and pruning specifically requires the WHERE clause to reference the partition key directly with values (or parameters) the planner can actually compare against those bounds.
Why can't partition pruning work with a WHERE clause like WHERE logdate >= CURRENT_TIMESTAMP, and what does that tell you about writing partition-friendly queries?
Partition pruning at plan time requires the comparison value to be known and fixed when the plan is built, and `CURRENT_TIMESTAMP` is not immutable, its value depends on when the query actually executes, not when it's planned, so the planner can't statically prove which partitions it will or won't match. (Pruning can still happen at execution time for genuinely parameterized values, like a join parameter from an outer query, just not for volatile functions like this one.) This means writing partition-friendly queries means filtering on the partition key with values the planner can actually reason about, a literal, a bound parameter, or an immutable expression, not a function whose result varies by when the query runs.
In a systemd unit, what is the practical difference between Type=simple and Type=forking, and why does that distinction matter for dependency ordering?
With Type=simple, systemd considers the unit started the moment the main process is forked off, it does not wait for the application to finish its own initialization, so anything depending on that unit might start before the service is actually ready to handle requests. Type=forking expects the traditional daemon pattern, the initial process forks and exits once it judges its own startup complete, so systemd marks the unit started as soon as that original process exits successfully, while the actual daemon keeps running as a separate, now-orphaned process. That only tracks the daemonization handoff, not genuine application readiness, a process can exit believing setup is done while it is still finishing initialization in the background, so Type=forking is a better signal than Type=simple but still not a readiness guarantee. Type=notify is the one that actually is readiness-safe: the service explicitly calls sd_notify to tell systemd exactly when it's ready, rather than systemd inferring readiness from process exit behavior at all.
Why does Linux split user information across /etc/passwd and /etc/shadow instead of keeping everything, including the password hash, in one file?
/etc/passwd has to be world-readable, ordinary tools and commands need to map UIDs to usernames and look up home directories or login shells for every user on the system. If password hashes lived there too, every local user could copy them out and run an offline cracking attempt. /etc/shadow holds the actual encrypted password (and related aging data) and is readable only by root, while /etc/passwd keeps a placeholder character (commonly `x`) in the password field, preserving the UID-lookup functionality everything else depends on without exposing anything crackable.
How to Build a Production-Ready Auto-Scaling Azure Web App with Modular Terraform (VMSS, Load Balancer & NAT Gateway)
From Basic Terraform to Production IaC: Building an Auto-Scaling Azure Web App with Modular Terraform.
Why does a real ring hash implementation give each host many positions on the ring instead of just one, and what problem would a single position per host cause?
A host thrown onto the ring at just one point can end up, purely by chance, owning a disproportionately large or small arc of the ring if the hash values happen to land unevenly, since with few points there's no averaging effect smoothing out the randomness. Assigning each host many positions on the ring, scaled by that host's intended weight, so a double-weight host gets roughly twice as many ring entries as a single-weight one, averages out that randomness across many smaller arcs per host, producing a much more even overall traffic distribution than a single coin-flip-like placement per host would.
When does semantic search (via embeddings) actually outperform traditional keyword search, and when might keyword search still win?
Semantic search wins when a query and the relevant document use different words for the same idea, "car won't start" matching a document about "vehicle fails to ignite", since embeddings compare meaning rather than literal tokens, which keyword search cannot do at all. Keyword search still wins, or at least remains necessary, for exact-match needs, an error code, a product SKU, a specific proper noun, where the literal string matters and a semantically "close" but textually different result is actually the wrong answer. This is why production search systems commonly combine both (hybrid search) rather than treating embeddings as a strict replacement for keyword matching.
At what point in the Terraform workflow are Sentinel (or similar policy-as-code) checks evaluated, and why does that timing matter?
Policy checks evaluate against the plan, the output of `terraform plan`, before `terraform apply` actually provisions anything, which means a policy violation blocks the run from proceeding to apply at all. Evaluating against the plan rather than the already-applied state is what makes this a preventive control instead of a detective one; the non-compliant resource is stopped before it exists, not flagged for cleanup afterward once it's already live and potentially already been exploited or has already incurred cost.
A consumer receives a message, starts processing, but crashes before deleting it. What happens to that message, and why is this actually the desired behavior?
Once the visibility timeout expires without the message being deleted, it automatically becomes visible again in the queue and can be picked up by the same or a different consumer for another processing attempt. This is deliberate, not a bug: the alternative (a crashed consumer's message vanishing permanently) would silently lose data, whereas reappearing after timeout guarantees eventual processing at the cost of a possible duplicate attempt, which is exactly the trade-off "at-least-once delivery" describes. Setting the visibility timeout too short causes premature, unnecessary reprocessing of messages still legitimately being worked on; too long delays legitimate retries after a real crash.
Why does wrapping several statements in `BEGIN`/`COMMIT` matter, even for a multi-step operation that's logically one action?
Without an explicit transaction, PostgreSQL commits each statement independently the moment it succeeds; if a multi-step operation (debit one account, credit another) fails halfway through, the database is left in an inconsistent, partially-applied state with no way to undo the completed step. Wrapping the statements in `BEGIN`/`COMMIT` groups them so they can be committed or rolled back together, but that alone isn't automatic protection against every failure mode: PostgreSQL only aborts a transaction on its own when a statement raises an actual SQL error, an `UPDATE` that runs successfully but matches zero rows (a mistyped account id, for instance) is not an error at all, and would otherwise be committed as if the transfer had actually happened. Making the transaction genuinely safe means checking that each `UPDATE` affected exactly the one row expected and issuing an explicit `ROLLBACK` if either check fails, only issuing `COMMIT` once both checks pass.
Python Data Structures
When to reach for a list, tuple, dict, or set based on what operations you actually need fast, and why choosing the wrong one is a common hidden performance bug.