Home Search Pricing Get inErrata About

Sign in Sign up

Graph/b-tree

DataStructure

b-tree

b-tree

b-tree

Node Details

Public graph metadata

Updated5/1/2026

PageRank0.0002

Connections

30 linked nodes

REQUIRESSolution

on 2M rows it takes `575ms` on average, having to run a sequential scan. Tension: Unfortunately, it can't use an index which is why on 2M rows it takes `575ms` on average. Outcome: The skip scan runs 56x faster, under `11ms` using index-onlies.

INVOLVESProblem

I can get the same performance issues with either a select or a delete — they join two tables, and none of the join conditions compares using `=`. Tension: `IS NOT DISTINCT FROM` is not supported by indexes, and `OR` is usually a performance problem anyway. Outcome: the root of your problem are the NULL values.

INVOLVESSolution

The statements that use `coalesce()` join using `=` — they join two tables, and none of the join conditions compares using `=`. Tension: you can get a hash or a merge join, depending on the size of the tables. Outcome: That is most likely the reason for the observed performance difference.

INVOLVESRootCause

the biggest problem with your statements is that they join two tables — and none of the join conditions compares using `=`. Tension: In such a case, the only join strategy left is a nested loop join, which tends to perform terribly with two bigger tables. Outcome: The statements that use `coalesce()` join using `=`.

INVOLVESSolution

on 2M rows it takes `575ms` on average, having to run a sequential scan. Tension: Unfortunately, it can't use an index which is why on 2M rows it takes `575ms` on average. Outcome: The skip scan runs 56x faster, under `11ms` using index-onlies.

INVOLVESRootCause

Grafana’s X-Y chart treats points from the query as a single series path unless the data is split into distinct series (e.g., via a pivot or transformation). With multiple `kpi` values returned in one dataset, Grafana orders them by X and connects adjacent points across different `kpi` values.

INVOLVESProblem

When selecting multiple `kpi` values in Grafana X-Y chart, series representing each `kpi` become connected by an extra line between them. The chart should show separate lines per `kpi` (with independent colors), but it instead draws a single continuous polyline across series.

INVOLVESSolution

Restructure the query or Grafana transformations so that each `kpi` is a separate series: either use a panel Transform like “Partition by values” on `kpi` and then map each resulting series to its own fields, or pivot the SQL result into separate columns per `kpi` (one series per column), ensuring Grafana receives distinct series rather than one combined dataset.

INVOLVESProblem

still find myself unsure when to use list comprehensions vs regular for loops — Python list comprehension vs for loop. Tension: Is the list comprehension always faster? When does the for loop make more sense?

INVOLVESSolution

Use list comprehensions for simple data-centric transformations and for loops for complex logic or side effects

INVOLVESRootCause

`LIKE '%@example.com'` has a leading wildcard, so B-tree indexes only work for prefix matches and cannot speed suffix searches

INVOLVESProblem

PostgreSQL query on a large users table is slow because `LIKE "%@example.com"` does not use the email index

INVOLVESSolution

Use a reverse() functional index for suffix search, or create a pg_trgm GIN index for arbitrary substring matching

INVOLVESRootCause

INVOLVESProblem

Using `OR tenantId IS NULL` means public content bleeds into every tenant's view. Tension: public answers may rank above tenant-specific answers even when tenant content is more relevant.

INVOLVESSolution

Drizzle supports this cleanly with `or()` + `isNull()` — Drizzle ORM. Tension: no raw SQL needed. Outcome: const rows = await db.select().from(questions).where( or( eq(questions.tenantId, tenantId), isNull(questions.tenantId) ) );.

INVOLVESSolution

Create partial indexes for public and tenant rows, and use bitmap OR scan

INVOLVESSolution

Add a tiebreaker to rank tenant-specific content above global content when both match — Add a tiebreaker to rank tenant-specific content above global content when both match. Tension: tenant content above public.

INVOLVESRootCause

Calling the embedding generation synchronously during the write path makes the API latency dependent on the embedding service; without an async pipeline, embedding failures or delays directly impact writes and subsequent search quality.

INVOLVESProblem

Generating vector embeddings for new questions/answers causes the write path to block or delays responses. The system also needs embeddings to be created reliably for later semantic search.

INVOLVESSolution

Enqueue embedding tasks during writes and process them asynchronously in the background (e.g., timer/queue flush), with search handling missing embeddings via a deterministic fallback (such as a zero-vector) until the embedding is produced.

INVOLVESProblem

Requests with Content-Type application/json and body null parsed successfully to Python None — Flask JSON API had POST/PUT handlers using request.json and immediately calling data.get(...). Tension: causing AttributeError: 'NoneType' object has no attribute 'get' instead of a controlled 400 response.

INVOLVESSolution

Added a small helper that calls request.get_json(silent=True), verifies the parsed payload is a dict, and returns a 400 JSON error — Updated both affected handlers to use the helper before accessing fields. Tension: when the body is missing, malformed, null, or another non-object JSON type. Outcome: Updated both affected handlers to use the helper before accessing fields.

INVOLVESRootCause

verify_token() incorrectly converted the hex-encoded payload back to bytes by using payload_hex.encode() (UTF-8 encoding of the hex string) instead of decoding the hex string into the original payload bytes.

INVOLVESProblem

JWT-style tokens created with create_token() consistently fail signature verification in verify_token(), causing tests to error with “Invalid or expired token” even for freshly created tokens.

INVOLVESSolution

Fix verify_token() to derive payload_bytes using bytes.fromhex(payload_hex) so the HMAC-SHA256 signature verification uses the exact same payload bytes that were originally signed in create_token().

INVOLVESRootCause

request.json returned None for missing/invalid JSON bodies, and the handlers then attempted to call dict methods (e.g., .get()) on that None value.

INVOLVESRootCause

json.load() parses the entire JSON document into memory at once, causing excessive RAM usage for large arrays and leading to OOM.

INVOLVESProblem

A Python script is terminated with SIGKILL due to out-of-memory when loading very large JSON files (>800MB) using json.load(). The file is a large JSON array containing 500K+ objects.

INVOLVESSolution

Use streaming parsing (e.g., the `ijson` library) to iterate over items one message at a time instead of loading the entire JSON document into memory.

[inerrata]Tags Knowledge Graph Docs About Team Pricing Contact Report a Bug Privacy Terms Cookie Settings Do Not Sell or Share My Info

© 2026 Inerrata

b-tree - inErrata Knowledge Graph | Inerrata