Ive tried to understand sharding for crypto purposes but every "crypto sharding" video display vague descriptions. I stumbled upon this by accident and it was great. Amazing work 🙌
As for the hashing, you probably will take hash only of a subset of columns of a record, most probably - only primary key, because changing any field of any record will result the hash to change also, which leads to data losses.
Thanks for this comment. Could you maybe clarify ? By data loss you mean the reverse proxy or whatever way of communicating with databases wont be able to know where to search for the information right ? however the information would still be here. Although it would be extremely computationally hard there would still be way to recover it right ? You could for example recalculate hashes for all the data and redistribute those that aren't belonging to the right database based on the sharding prefix ?
Hi, Anton! How can I search by the field that is not shard key? I need to go thru all the shards? And what if I need to scale it up or down (change shards number)?
Let's say I'm using a shard-nothing architecture, now let's say there's a relationship between customers table, payments table and orders table. Customers and orders tables are linked by the foreign key customers->id ~ orders->customer_id Orders table and payments table has the foreign key orders->id ~ payments->order_id Now how would you shard this database? You can't use a single shard key, because both customer_id and order_id are important that ensure all the related rows are in a single shard. So how would you solve this problem?
if a shard fails you mentioned it can be restored from other shard, but each shard has unique set of values. Could you help me understand how this restores
not sure if i follow, some distributed systems have not only sharding but also replication. so you have a primary shard and a replica for that shard. so if it fails, the replica becomes primary, but it's very specific to each database
READ TWICE, tried explaining my best!! A physical shard contains several primary logical shards and several replica logical shards. Primary and replica logical shards are always stored on different physical shards to ensure fault tolerance. If a physical shard goes down due to an issue (e.g., a server failure), there is no need to worry about the replicas, as their corresponding primaries are stored on other physical shards. However, the primary logical shards on the affected physical shard require attention. The Shard Manager must retrieve these primary shards from their replicas stored on other physical shards to maintain system functionality.
Privet, spasibo! At this point, I can offer one-on-one sessions. I ask that you send me any questions before the session so that I can prepare some examples, and we can go through them during the meeting. I charge $100 per 1 hour session. If you are interested, pls send me an email.
Great content, as always! A bit off-topic, but I wanted to ask: My OKX wallet holds some USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). Could you explain how to move them to Binance?
thanks, but i have a question if i use range-based sharding and conside 4 shard what happen if i want to convert to 40 shard? what happen for previous data, and new data [ first i have 3 shard 1(a-h) 2(i-p) 3(q,z)) now need to make it 40.
you said sharding have unique data sets if one sharding not respond then other sharding response you but if customer search record and that record will be in sharding 1 . After that sharding 1 will not respond then what we have to show for customer
Fortunately I’ve been able to get by with two read databases and a write by using table partitioning up until this point. Hopefully I don’t have to tackle sharding any time soon! Great video and thanks for sharing
Thanks for the forecast! 📊 Need some advice: 🙏 I have these words 🤨. (behave today finger ski upon boy assault summer exhaust beauty stereo over). Can someone explain what this is? 😅
So, sharding is a pain in the ass and requires a lot of configuration, analytics and also business logic to manage shards in an respectable way. This is also why NoSQL Databases come in handy as they can scale better horizontaly without this extensive configuration activities you have with traditional SQL databases. But to be said SQL Databases will probably cover 90% of all usecases anyway without you getting into sharding.
🔴 - To support my channel, I’d like to offer Mentorship/On-the-Job Support/Consulting - me@antonputra.com
Hard to believe this top-notch content has very less views. Thanks a lot, Anton!!
❤️
probably coz it basically has no volumn
Ive tried to understand sharding for crypto purposes but every "crypto sharding" video display vague descriptions. I stumbled upon this by accident and it was great. Amazing work 🙌
thanks!
No Nonsense, direct to point, covering all cases.
Well-compiled video!
thank you!
As for the hashing, you probably will take hash only of a subset of columns of a record, most probably - only primary key, because changing any field of any record will result the hash to change also, which leads to data losses.
Thanks for this comment. Could you maybe clarify ? By data loss you mean the reverse proxy or whatever way of communicating with databases wont be able to know where to search for the information right ? however the information would still be here. Although it would be extremely computationally hard there would still be way to recover it right ? You could for example recalculate hashes for all the data and redistribute those that aren't belonging to the right database based on the sharding prefix ?
@@tonnytrumpet734 oh yes, data will still be there, but it basically will become unusable. It's like creating yourself problems to solve.
Very helpful. I just discovered sharding not long ago in my new position. I didn't understand a thing until your video.
thank you!
Thank you sir for detail explaination of database sharding. We hope a practical handson of Database sharding will publish soon.
Nice, exhaustive and short video considering it covers a lot.
Great Explanation....Thanks for the efforts
Thanks!
Great video! Straight to the point and very explicit. Thanks
Very helpful. Very confused and to the point! I hope your colleagues who do technical videos would follow your framework. 👍🏾👍🏾👍🏾
Great Video, as always, Anton!
Thanks, if you think anything can be improved, please let me know!
Great Questions on sharding and good explaination followed by....
This should be the approach for great teaching
thanks!
It has been such a lovely recap. Thanks a lot for useful content
thanks!
Superb explanation, and never strayed off topic.
Great video! Always an important topic when we think about scale our systems.
Thank you!
Great video Anton.
Thank you!
Hi, Anton! How can I search by the field that is not shard key? I need to go thru all the shards?
And what if I need to scale it up or down (change shards number)?
Great video. Clear & easy to understand.
Thanks Ab!
Thank you so much for this clear, insightful explanation of Database Sharding.
Thank you, Charles!
please make tutorial how to sharding database in vps like hetzner or contabo or ovh, since my budget not suitable for at the moment using aws now
Let's say I'm using a shard-nothing architecture, now let's say there's a relationship between customers table, payments table and orders table.
Customers and orders tables are linked by the foreign key customers->id ~ orders->customer_id
Orders table and payments table has the foreign key
orders->id ~ payments->order_id
Now how would you shard this database? You can't use a single shard key, because both customer_id and order_id are important that ensure all the related rows are in a single shard.
So how would you solve this problem?
if a shard fails you mentioned it can be restored from other shard, but each shard has unique set of values. Could you help me understand how this restores
not sure if i follow, some distributed systems have not only sharding but also replication. so you have a primary shard and a replica for that shard. so if it fails, the replica becomes primary, but it's very specific to each database
@@AntonPutra got it, so each shard will have multiple replica's for failover?
READ TWICE, tried explaining my best!!
A physical shard contains several primary logical shards and several replica logical shards. Primary and replica logical shards are always stored on different physical shards to ensure fault tolerance.
If a physical shard goes down due to an issue (e.g., a server failure), there is no need to worry about the replicas, as their corresponding primaries are stored on other physical shards. However, the primary logical shards on the affected physical shard require attention. The Shard Manager must retrieve these primary shards from their replicas stored on other physical shards to maintain system functionality.
Классное видео, спасибо. Какая стоимость ваших услуг? Нужна консультация для Homelab.
Privet, spasibo! At this point, I can offer one-on-one sessions. I ask that you send me any questions before the session so that I can prepare some examples, and we can go through them during the meeting. I charge $100 per 1 hour session. If you are interested, pls send me an email.
Great content, as always! A bit off-topic, but I wanted to ask: My OKX wallet holds some USDT, and I have the seed phrase. (alarm fetch churn bridge exercise tape speak race clerk couch crater letter). Could you explain how to move them to Binance?
thanks!
Truly awesome and simple to learn!!! Thank you!!!
Thank you for the feedback!
Great explaination! Thanks
thank you!
Really very well explained, thank you very much.
thanks!
So how to solve hotspot problem?
Awesome video! thanks for explaining it
Thanks!
Excellent tutorial
Thank you!
Excellent described
Thanks!
I never expected Elijah Wood to teach me Data Science. Well done, Elijah.
😊
thank you for these explanations
thanks for visiting
sir thanks for the video, what do you use for editing, its really good.
thanks, adobe suite
Can I do sharding in WordPress database?
WordPress uses a MySQL database under the hood. Take a look at Vitess.
How do you create the animation for your videos? They are so cool!!!!! 💪🏼
Thanks! I use adobe suite.
thanks, but i have a question if i use range-based sharding and conside 4 shard what happen if i want to convert to 40 shard? what happen for previous data, and new data [ first i have 3 shard 1(a-h) 2(i-p) 3(q,z)) now need to make it 40.
If you shard manually at the application level, you need to write logic to rebalance it yourself. It's easier to use built-in mechanisms for sharding.
@@AntonPutra thanks
Good explanation
thanks!
Very interesting! Thanks! 👍
welcome! my pleasure
you said sharding have unique data sets if one sharding not respond then other sharding response you but if customer search record and that record will be in sharding 1 . After that sharding 1 will not respond then what we have to show for customer
Thank you much appreciated 👍
thanks :)
Great video, thank you
thanks!
Fortunately I’ve been able to get by with two read databases and a write by using table partitioning up until this point. Hopefully I don’t have to tackle sharding any time soon! Great video and thanks for sharing
Thanks! There a lot of distributed databases based on postgres that can shard for you.
Excellent! What tool do you use to do animations?
Thanks Adobe suite
Short,nice,clear
Good video
and here is that one video which clarifies it simply, thank you. you made it easy like drinking milk.
thanks :)
Thanks for the forecast! 📊 Need some advice: 🙏 I have these words 🤨. (behave today finger ski upon boy assault summer exhaust beauty stereo over). Can someone explain what this is? 😅
?
thanks a lot for explaining
my pleasure!
thanks bro
tutorial like south Indian suspense thriller movies , cant blink your eyes
i know, getting used to :)
I bet this guy works for the mafia
😂
your example confusing between database shard and table partitioning
range-base sharding is about one table sharding not about database sharding.
Noted, will improve
I think 99.9% use case are served fine by a monolith database server. Heck even stack overflow is fully powered on a single server
Yes. Still, it's useful to have a knowledge of whatever jargons uppermamagements are throwing at you.
For personal projects, sure, but in the enterprise, you frequently have to deal with sharding.
So, sharding is a pain in the ass and requires a lot of configuration, analytics and also business logic to manage shards in an respectable way.
This is also why NoSQL Databases come in handy as they can scale better horizontaly without this extensive configuration activities you have with traditional SQL databases.
But to be said SQL Databases will probably cover 90% of all usecases anyway without you getting into sharding.
Nowadays, once a year, I see a new distributed database based on PostgreSQL come out, lol.
great video! start subscribing now
Thank you!
w video
thanks
Bro why are you in such hurry ?
🤷
Feels like a bot is reading the script. Good content, but please act it out a bit.
thanks for the feedback