Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
shahargl
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
We used Datadog to save $17.5M annually
(datadoghq.com)
2 points
by
shahargl
2y ago
|
0 comments
32.
▲
The One Million Checkboxes Mystery
(twitter.com)
2 points
by
shahargl
2y ago
|
1 comments
33.
▲
by
shahargl
2y ago
How does it play with Keycloak's Authorization Services? Or in other words - why should I choose permify over Keycloak? https://www.keycloak.org/docs/latest/authorization_services/...
34.
▲
by
shahargl
2y ago
https://github.com/keephq/keep (disclaimer - i'm the maintainer)
35.
▲
I hacked Microsoft's AI bot for healthcare on a Friday night
(twitter.com)
3 points
by
shahargl
2y ago
|
0 comments
36.
▲
Jewish Problems: Problems Designed to Prevent Jews Getting a Passing Grade
(arxiv.org)
5 points
by
shahargl
2y ago
|
1 comments
37.
▲
Grafana Labs Observability Survey 2024
(grafana.com)
20 points
by
shahargl
3y ago
|
17 comments
38.
▲
Gitlab Acquisitions Handbook
(handbook.gitlab.com)
3 points
by
shahargl
3y ago
|
0 comments
39.
▲
Discovering AIOps
(apmdigest.com)
2 points
by
shahargl
3y ago
|
0 comments
40.
▲
Three new on-call providers emerge in a week. What gives?
(bobbytables.io)
1 points
by
shahargl
3y ago
|
0 comments
41.
▲
Nukemap – Atomic Bomb Simulator
(nuclearsecrecy.com)
5 points
by
shahargl
3y ago
|
0 comments
42.
▲
Snowflake Python connector lack of async support
(github.com)
2 points
by
shahargl
3y ago
|
0 comments
43.
▲
OpenSSH 9.6
(openssh.com)
1 points
by
shahargl
3y ago
|
0 comments
44.
▲
2023 IT Ops VendorScape [pdf]
(bigpanda.io)
2 points
by
shahargl
3y ago
|
0 comments
45.
▲
Observability for LLM's with Opentelemetry
(traceloop.com)
4 points
by
shahargl
3y ago
|
0 comments
46.
▲
by
shahargl
3y ago
how is it different from Traceloop and openllmetry ( https://github.com/traceloop/openllmetry )?
47.
▲
The ticking time bomb of observability expectations
(davidkcaudill.medium.com)
2 points
by
shahargl
3y ago
|
0 comments
48.
▲
by
shahargl
3y ago
one of the concepts we are trying to push is exactly that - my cofounder actually wrote a blog post about it - https://news.ycombinator.com/item?id=38086082
49.
▲
by
shahargl
3y ago
So the goal is to have unified interface for every tool you’ve just mentioned
50.
▲
by
shahargl
3y ago
valid feedback. If a company works only with Datadog it probably doesn't need a unified API for alerts, but what we see is that many companies use more than one. monitoring tool (or even other tools that generates alerts but arent clas
51.
▲
by
shahargl
3y ago
the roadmap is available at https://github.com/keephq/keep/issues feel free to create ticket if anything missing
52.
▲
Single alerts API for all monitoring tools
(keephq.dev)
39 points
by
shahargl
3y ago
|
13 comments
53.
▲
Ask HN: Good blog posts about react/TypeScript build process
1 points
by
shahargl
3y ago
|
1 comments
54.
▲
Northrop Grumman B-21 Raider
(en.wikipedia.org)
1 points
by
shahargl
3y ago
|
0 comments
55.
▲
Elastic Observability 8.11
(elastic.co)
1 points
by
shahargl
3y ago
|
0 comments
56.
▲
by
shahargl
3y ago
100%
57.
▲
by
shahargl
3y ago
Yet again, that’s about the data, I think the blog post focuses on the applicative aspect
58.
▲
by
shahargl
3y ago
But that’s for the data part (otel) and not the applicative part (alerts, dashboards, etc)
59.
▲
by
shahargl
3y ago
Cool! Never heard of it and definitely gonna check it.
60.
▲
by
shahargl
3y ago
Wdym by moving people?
More ›