Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mr-karan
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
91.
▲
by
mr-karan
4y ago
I am getting `Access denied Error code 1020` while trying to access https://techjobsforgood.com/ from AS55836 in India.
92.
▲
by
mr-karan
4y ago
Amazing, thanks for your response. I'll explore Docusaurus 2.0 :) Congrats on the release.
93.
▲
by
mr-karan
4y ago
I'm so happy with Material Mkdocs[1]. I use Obsidian to write a bunch of notes and then combine that with a bunch of mkdocs plugins to generate a static site out of it [2]. What does Docusaurus offer more, if anyone who's used bot
94.
▲
by
mr-karan
4y ago
I think LogQL[1] is a pretty good alternative to Lucene search syntax. It feels more natural and human friendly for searching logs. [1]: https://grafana.com/docs/loki/latest/logql/
95.
▲
by
mr-karan
4y ago
This is quite a cool project! Congrats on the public release.
96.
▲
by
mr-karan
4y ago
Hey Armin. That's very nice to hear that reduction is something on your roadmap. Pre Snuba, the number of external services were fairly limited but I understand as Sentry got more mature and added a bunch of extra services than "j
97.
▲
by
mr-karan
4y ago
This is so true. Sentry's architecture[1] is insanely complex. It runs some ~30 containers on the instance and all you've to pray is that their tooling[2] for `docker-compose` _just works_. [1]: https://develop.sentry.d
98.
▲
by
mr-karan
4y ago
Honestly, I like to keep the logger as decoupled and minimalist as possible. There’s https://vector.dev which can basically ship your logs from source A to destination B (supports tonnes of usecases). Separation of concerns make
99.
▲
by
mr-karan
4y ago
Yup, I use a self hosted version of [1] [1]: https://github.com/sissbruecker/linkding
100.
▲
by
mr-karan
4y ago
They've just updated: https://www.cloudflarestatus.com/incidents/xvs51y9qs9dj
101.
▲
by
mr-karan
4y ago
Also shoutout to https://github.com/sissbruecker/linkding which is a nice self hosted bookmarks service that I've been using over the past one year or so.
102.
▲
by
mr-karan
4y ago
Things were slow today morning (3 hours ago), from India.
103.
▲
Understanding Networking in Nomad
(mrkaran.dev)
7 points
by
mr-karan
4y ago
|
0 comments
104.
▲
by
mr-karan
5y ago
Noted, that makes sense. Thanks!
105.
▲
by
mr-karan
5y ago
I did mention that it's the OP's server which I presume isn't in an autoscale group. Even then, saving a few MBs in image size is the devops parlance of early optimisation. There's so much that happens in an Autoscale gr
106.
▲
by
mr-karan
5y ago
While this is remarkably a good hack and I did learn quite a bit after reading the post, I'm simply curious about the motivation behind it? A docker image even if it's a few MBs with Caddy/NGINX should ideally be just pulled
107.
▲
Nomad 1.3 Beta Adds Native Service Discovery and Edge Workload Support
(hashicorp.com)
5 points
by
mr-karan
5y ago
|
0 comments
108.
▲
by
mr-karan
5y ago
Folks looking for a solution to storing Prometheus metrics from multiple places, definitely consider exploring Victoriametrics. I'm running a single Victoriametrics instance which has 230bn metrics, consuming ~4GB of memory and barely
109.
▲
by
mr-karan
5y ago
[rich]( https://news.ycombinator.com/item?id=29945848 ) was recently discussed as well.
110.
▲
by
mr-karan
5y ago
FWIW, Logseq[1] can do this outliner stuff. [1]: https://logseq.github.io/#/page/Contents
111.
▲
Nomad Tips and Tricks
(danielabaron.me)
2 points
by
mr-karan
5y ago
|
0 comments
112.
▲
Ask HN: Project ideas for learning distributed systems
2 points
by
mr-karan
5y ago
|
7 comments
113.
▲
by
mr-karan
5y ago
Which is nice, except Obsidian isn't FOSS (the parent comment was about free/open source editors).
114.
▲
by
mr-karan
5y ago
Lovely! Thanks for buliding it.
115.
▲
by
mr-karan
5y ago
AS6453 is having a mega outage according to https://www.thousandeyes.com/outages/ . Lot of services for users in India were inaccessible for ~1h or so.
116.
▲
by
mr-karan
5y ago
For a very long time, I've been using shynet [0] on my personal blog. I've not played around the alternatives but there are quite a few: - Plausible: https://plausible.io/ - umami: https://umami.is
117.
▲
by
mr-karan
5y ago
Lost you at "Typical FOSS dog shit" I've used Joplin and while it has some rare rough edges (but then what hasn't) but my comment isn't about Joplin per-se. You really sound super entitled when you demand for what
118.
▲
Using ClickHouse Keeper for Replication
(mrkaran.dev)
3 points
by
mr-karan
5y ago
|
0 comments
119.
▲
by
mr-karan
5y ago
The backing table in OP's case is just 2 columns, timestamp and a `String` column representing JSON blob. Can't do much there.
120.
▲
by
mr-karan
5y ago
Fwiw, Clickhouse Keeper _can_ be run as an external daemon if you wish so. But it's packaged with the server binary itself, so once you add <keeper_config> to all your server nodes, you're good to go, without running anythin
More ›