Forwarded from Ṭech_Ṣteps Hub (Kalkidan)
Sometimes, all we need is someone who lifts our spirits, believes in us, and reminds us that no matter what we're facing, we are capable of overcoming it.
Ṭech_Ṣteps Hub
Sometimes, all we need is someone who lifts our spirits, believes in us, and reminds us that no matter what we're facing, we are capable of overcoming it.
Idk about y’all, but that someone is my bro to me. He’s not my sub so I can appreciate him freely (habsha yekoral sednek begltse so i choose to hide)
Is it possible before the academic year starts me to hit 300subs 😁 so I can flex on my Gigi crush?
💯2😁1
dag.shell
so I can flex on my Gigi crush?
This media is not supported in your browser
VIEW IN TELEGRAM
Birhan Nega
#ከአስር_አመታት_ተሞክሮየ
#ክፍል-አንድ
ምንም ይሁን አንድ ጉዞ ስትጀምሩ አብሯችሁ የሚጓዝ ሰው ሲኖር ፍጥነታችሁን ያሻሽለዋልዋል። ሰው አዋዋዩን ይመስላል ይባላል አይደል? ሯጮች እንኳን ሪከርድ ማሻሻል ሲፈልጉ pacemaker ይጠቀማሉ።
#ክፍል-አንድ
ምንም ይሁን አንድ ጉዞ ስትጀምሩ አብሯችሁ የሚጓዝ ሰው ሲኖር ፍጥነታችሁን ያሻሽለዋልዋል። ሰው አዋዋዩን ይመስላል ይባላል አይደል? ሯጮች እንኳን ሪከርድ ማሻሻል ሲፈልጉ pacemaker ይጠቀማሉ።
such important point gin ካልተገኝስ 🤧 u got walk the road by ur self ig
👀lesson of the day(of backend guys)
The Failure-tolerant systems
Let’s say you have a Telegram bot. Telegram sends an event to your server through a webhook, and your worker is responsible for processing that event.
Now imagine your worker receives the event, but something fails while processing it.
If your system simply tells Telegram, “Yep, I got it ✅” before the processing is actually completed, Telegram assumes everything went fine. Your worker failed, but the event is already considered handled.
Now you’ve lost the event 🤧😁
Sure, Telegram may keep updates available for some time, but your application logic shouldn’t depend on that as a recovery mechanism. Your system should be designed so that a temporary worker failure doesn’t mean permanent data loss.
That’s where queues come in.
Instead of:
"Webhook → Worker → if worker fails, event is gone "💀
You can do:
"Webhook → Queue → Worker → Process"
The queue acts as a buffer between receiving an event and processing it. If the worker crashes, fails, or needs to restart, the event can remain in the queue and be processed again.
That’s basically the idea behind building a failure-tolerant system: assume things will fail, and design the system so failures are recoverable.
For my current setup, I’m using Supabase for pretty much everything so they have Supabase Queues (via some pg db extension )
As they say pg db for all since supba is on top of it supba for all😁
The Failure-tolerant systems
Let’s say you have a Telegram bot. Telegram sends an event to your server through a webhook, and your worker is responsible for processing that event.
Now imagine your worker receives the event, but something fails while processing it.
If your system simply tells Telegram, “Yep, I got it ✅” before the processing is actually completed, Telegram assumes everything went fine. Your worker failed, but the event is already considered handled.
Now you’ve lost the event 🤧😁
Sure, Telegram may keep updates available for some time, but your application logic shouldn’t depend on that as a recovery mechanism. Your system should be designed so that a temporary worker failure doesn’t mean permanent data loss.
That’s where queues come in.
Instead of:
"Webhook → Worker → if worker fails, event is gone "💀
You can do:
"Webhook → Queue → Worker → Process"
The queue acts as a buffer between receiving an event and processing it. If the worker crashes, fails, or needs to restart, the event can remain in the queue and be processed again.
That’s basically the idea behind building a failure-tolerant system: assume things will fail, and design the system so failures are recoverable.
For my current setup, I’m using Supabase for pretty much everything so they have Supabase Queues (via some pg db extension )
As they say pg db for all since supba is on top of it supba for all😁
If ur looking for project idea here it's
As a certified bus user 🤧😁
what if there was an AI-powered app that could help blind people identify which bus is standing in front of them? I have seen them struggle a lot,they miss a lot of buses if there is no one around them to ask
As a certified bus user 🤧😁
what if there was an AI-powered app that could help blind people identify which bus is standing in front of them? I have seen them struggle a lot,they miss a lot of buses if there is no one around them to ask
🔥2
dag.shell
As a certified bus user 🤧😁
This media is not supported in your browser
VIEW IN TELEGRAM
Soset Hub
Pulling all nighters @ abrhot with @dag_shell
@MickyCodes ❌
Mickey (the Disney character) stickers ✅
Anyway
@dag_shell x @soset_ge in abechu building
Mickey (the Disney character) stickers ✅
Anyway
@dag_shell x @soset_ge in abechu building
😁5
dag.shell
@MickyCodes ❌ Mickey (the Disney character) stickers ✅ Anyway @dag_shell x @soset_ge in abechu building
Afternoon(morning to me) lock-in session with ma little one sis
🔥4❤1
dag.shell
Afternoon(morning to me) lock-in session with ma little one sis
Owww look what she got out of the book z GOAT 🐐
dag.shell
Owww look what she got out of the book z GOAT 🐐
yo if you're seeing this @tamas_archive it's ready to be archived i mean for sale 👀😁
❤1
dag.shell
npm uninstall *rase-metat* --from=*me* 🤧
Egnam last SMA of the year enargi enda 👀
dag.shell
Egnam last SMA of the year enargi enda 👀
how this could be a solution to my *rase-metat* 🤧
This media is not supported in your browser
VIEW IN TELEGRAM
Me after i uninstall that rase-metat with cold 🥶 shower on morning bezi berdi 🤧
😁5