Problemsby GlyphObservability and logging costs outgrow the value
8 posts from 8 people8d activeSteady, +1 this week
Posts per day
The posts behind it
Expose Delta Lake log statistics to the query optimizer
With `use_iceberg_manifest_statistics`, join reordering and the Cascades optimizer estimate Iceberg reads from their manifest files: rows after pruning and deletes, plus min/max, NULL fraction and number of distinct values of the columns. Delta Lake reads still get no estimate....
Datadog: We're paying the bill, so why the hell can't we access our cost data?
...I'm trying to understand where the money goes. Why is accessing our own cost data such a nightmare?. Has anyone dealt with anything similar?.
Docs: self-hosting cacheHandler example is an unbounded in-memory cache and disables the default LRU
### What is the documentation issue? The `cache-handler.js` example in [Self-Hosting → Configuring Caching](https://nextjs.org/docs/app/guides/self-hosting#configuring-caching) stores entries in a module-level `Map` with no size limit, no TTL, and no eviction. The paired...
[Usage]: How to reduce the number of Prometheus metrics exposed by vLLM?
### Your current environment ### What is the problem? I deploy vLLM on Kubernetes and scrape the `/metrics` endpoint with a Prometheus Operator `ServiceMonitor`. The endpoint exposes a very large number of metrics/time series (about [N] series per pod). This is a concern for us...
Panic in row estimation for group_by/unique over an empty parquet scan under a join
### Checks - [x] I have checked that this issue has not already been reported. - [x] I have confirmed this bug exists on the [latest version](https://pypi.org/project/polars/) of Polars. (broken in 2.0.0-rc.2, fine on 1.44.2.) ### Reproducible example ### Log output ### Issue...
RFC Elasticsearch Table Engine
### Company or project name ClickHouse ### Use case To be able to query Elasticsearch form within ClickHouse ### Describe the solution you'd like # Arguments for `Elasticsearch` 1. Positional arguments, where some of them have `key = value` format 2. Named collection with index...
langchain-typesafe: TypeSafeClassifier sends requests when questions is empty
### Submission checklist - [x] This is a bug, not a usage question. - [x] I added a clear and descriptive title that summarizes this issue. - [x] I used the GitHub search to find a similar question and didn't find it. - [x] I am sure that this is a bug in LangChain rather than...
Arrow scan: estimated cardinality is always 1, so `ORDER BY ... LIMIT` loses `TOP_N`
### What happens? For a registered Arrow table, the estimated cardinality that reaches the optimizer is always 1, and it arrives flagged as a known estimate rather than as absent. This degrades optimizer decisions (ie: TopN selection) for arrow sources. ## Impact The estimate...
Companies and products named
About this problem
Evidence
8 posts from 8 people in 7 places, about 1 a week over 56 days. Mostly on ClickHouse/ClickHouse, langchain-ai/langchain, duckdb/duckdb. Tools named alongside: GitHub, ClickHouse, Kubernetes, LangChain.
Frustration Frustration 1 of 3· Seen on GitHub issues, Reddit
How it was grouped
Posts that state a pain and match the "observability-cost" rule. First post Sep 11, 2026, latest Nov 5, 2026. Corroborated: 3 or more posts from 2 or more people or places. Method
The same problem elsewhere
ClickHouse costs outgrow the value it gives
11 posts from 3 people on GitHub issues.
DuckDB costs outgrow the value it gives
4 posts from 4 people on GitHub issues.
Polars costs outgrow the value it gives
2 posts from 2 people on GitHub issues. Unverified.
NVIDIA costs outgrow the value it gives
2 posts from 2 people on GitHub issues. Unverified.
dbt costs outgrow the value it gives
2 posts from 1 person on GitHub issues. Unverified.
History