Very important video. Kafka, Asynchronous messaging is forming the main part of backend Java developer interviews these days. Thanks team. Keep such content coming !
Thanks a lot for helping us to prepare for the interview.. Also the baby from the background is also trying to teach us something..it is sad we can't understand 😅
I was asked recently in morgan stanley interview how consumer handles messages coming in sequence if the consumer fails. Wish i saw this video before. Useful content ❤
Kafka Replicator. You need to Configure like this ---> NewTopic createTopic() { return TopicBuilder.name("product-created-events-topic") .partitions(3) .replicas(3) .configs(Map.of("min.insync.replicas","2")) .build();
Thanks for great videos. Please create a video on kafka rebalancing, how to clear lag in kafka, how to decide how many partitions to chose for a kafka topic.
Hi, Thanks to Code Decode team, Can you please clarify the below points; 1. Can producer publish data on followers also or only on leader? 2. Will offset be removed after consumer consumed, if it will not remove till some time,then data will be huge data in the partition, then what is the limit of partitions? 3. Can multiple consumers consume the same offset? 4. how different consumers knows from which offset they need to start consuming? 5. Can multiple consumers from the same group or different groups consume same offset? 6. how to handle consumer lag any case of consumer fail to consume the data?
1. Only producer and followers will replicate data 2. Once offset is consumed, it’s consumed for particular consumer group so message can still be consumed again. There is retention policies, after that data will be deleted permanently 3. Yes, as long as consumers belong to different group 4. No, as partition data us read by specific consumer of group only 5. Fine tune buffer size, rebalance consumers or topic partition etc
Thanks to the Code Decide team, it's great I have been following ur classes for quite some time really appreciated. In this particular video, (1)What are the reasons that the leader broker dies ? (2) How the Leader Election does -- what's the Behind Algorithm and how this ISR works when data is produced in a broker/partition ? (3) Why Exactly Three 3 copies of Follower Brokers with the same set of data in Partitions ? (4) Do We Now have a successful Kafka running version Without ZOOKEEPER Dependency ?? which Version is it ?? How successful this new version in market is now ?? Please reply with your insights and answers. Thanks keep doing this nice work and knowledge sharing for a lot of aspirants to become successful IT careers like us. 👍🙏😊
1) broker is nothing but a server, which we start with a command. Now it can stop or crash any time just like our normal servers. 2) using epoch number. The most recently and most updated isr / broker is choosed az as new leader
How to handle the scenario where a consumer gets delayed reading a message, for example- Consumer B reads "booking successful" message before and a Consumer A could read "payment failure" message ? When both messages are from different topics
I am not sure if I understood your questions completely how kafka will manage if data got consumed in different order like order placed is consumed without payment message is received. I will suggest there will be two different microservices one will consume data from payment topic and once confirmed payment is done it will publish data to a different topic from where booking service will book the order Now second scenario if there is single microservice consuming from a single topic then it's code responsibility to check payment status first then go for booking. Please correct me if I am wrong
Your presentation is very good. 👍 There are so many videos on Udemy which don’t follow minimum standards. I really liked the way you explained the concept. I am not sure if you have your training courses on udemy as well. If not you can try that and you will be a star teacher over there. Just one suggestion though, don’t repeat a point of explanation multiple times, to make your video crisp and clear.
Hi please clarify that how can u ensure that Kafka consumes the messages in the sequential? 2. Will publisher publish duplicate messages? 3. How to retailn failed messages from the dlq within 7 days?
Nice vedio on Kafka, but there is a small mistake. You said both Producer writes and Consumer reads from Leader partition, which is incorrect. As this would stop the concept of parallelism when there are more than one consumer, and both trys to read from leader in your case. Generally, the leader status of the partition does not have any bearing on the consumer. The leader status only applies to the producers and the broker, so the consumer is fine to read all its data from both the follower Partition and the leader Partition.
"The leader handles all read and write requests for the partition while the followers passively replicate the leader" As per the documentation of Kafka itself. kafka.apache.org/081/documentation.html To answer your question on parallelism, "Each server acts as a leader for some of its partitions and a follower for others so load is well balanced within the cluster." Again as per the documentation 😊😊
@@CodeDecode Thanks.. I could'nt able locate the info provided in the documentation.. But for simplicity, lets say we only have one topic, three brokers and 3 consumers belongs to different consumer group. Now per the documentation only one Broker can have a leader for this one Topic and rest 2 are followers. (I assume we both agree on this) In this case, lets say reads happens only from the leader, then all three consumer would read only from the leader for this topic, which stops parallelism right. So ideally for Kafka to support parallelism here, one broker is assigned per consumer, so one of the consumer may read from leader , but the other 2 has to read from the follower. Otherwise in your case, all consumer will read from leader, which is from same broker, which distrupts parallelism. so here's a more accurate representation: "The leader handles all write requests for the partition, actively replicating the data to its followers. While the leader is responsible for coordinating writes, both leaders and followers can handle read requests. Consumers can read data from both the leader and followers, providing fault tolerance and parallelism." please correct me if my understanding is not correct .. still learning, and thanks for the reply..
Your explanation seems reasonable to me too but am still confused as documentation is contradicting it. Give me some time. Will try to connect to open source team. Either they will revise the documentation or else our doubt will be cleared
Very important video. Kafka, Asynchronous messaging is forming the main part of backend Java developer interviews these days. Thanks team. Keep such content coming !
Thanks Sayan 😊
Working in Kafka for the past 2 years without knowing much but just configuring topics. Now u made me understand the whole. Thanks!
😊😊👍
Awesome video. Saved my 2-3 hours of reading. Waiting for part-2
😊😊
Awesome content. Thanks for sharing knowledge.
You are welcome
next level of explanation . want more videos on this👌👌
Sure we will create more on this
I'm sure this channel will grow rapidly and extensively. Content quality is too good.
You are the Best 🙏🙏
😊😊 Thanks 👍😊
Thank you for this video. Eagerly awaiting for the Part-2
Sure we will upload it soon
Wonderful video. Learn a lot in 30mins about kafka.
😊😊
great video ,waiting for part two
Sure, we will create it soon
Very beautiful explanation.
Thanks
Excellent Video , Thanks for the detailed notes
You’re welcome
Very useful. Please upload this video continuation too. Thanks a lot!
Thanks a lot for helping us to prepare for the interview..
Also the baby from the background is also trying to teach us something..it is sad we can't understand 😅
🙊🙊🙊🙊 sorry for the inconvenience❤️ And thanks for understanding ,😘
Thanks for the video. Very informative.
You're welcome!
I was asked recently in morgan stanley interview how consumer handles messages coming in sequence if the consumer fails. Wish i saw this video before. Useful content ❤
Thanks 😊👍
Kafka Replicator. You need to Configure like this ---> NewTopic createTopic() {
return TopicBuilder.name("product-created-events-topic")
.partitions(3)
.replicas(3)
.configs(Map.of("min.insync.replicas","2"))
.build();
You have made video with love, I like it
Thanks
Very informative video on kafka ..always referring these videos with brilliant teacher to have us..thanks for come up with such videos.
Thanks for great videos. Please create a video on kafka rebalancing, how to clear lag in kafka, how to decide how many partitions to chose for a kafka topic.
Noted👍
Hi, Thanks to Code Decode team, Can you please clarify the below points;
1. Can producer publish data on followers also or only on leader?
2. Will offset be removed after consumer consumed, if it will not remove till some time,then data will be huge data in the partition, then what is the limit of partitions?
3. Can multiple consumers consume the same offset?
4. how different consumers knows from which offset they need to start consuming?
5. Can multiple consumers from the same group or different groups consume same offset?
6. how to handle consumer lag any case of consumer fail to consume the data?
1. Only producer and followers will replicate data
2. Once offset is consumed, it’s consumed for particular consumer group so message can still be consumed again. There is retention policies, after that data will be deleted permanently
3. Yes, as long as consumers belong to different group
4. No, as partition data us read by specific consumer of group only
5. Fine tune buffer size, rebalance consumers or topic partition etc
very well explained
Thanks !!
I was waiting for this 😮 😊 please continue thanks for the knowledge 🙏
Sure thanks 😊😊
Thanks to the Code Decide team, it's great I have been following ur classes for quite some time really appreciated. In this particular video,
(1)What are the reasons that the leader broker dies ?
(2) How the Leader Election does -- what's the Behind Algorithm and how this ISR works when data is produced in a broker/partition ?
(3) Why Exactly Three 3 copies of Follower Brokers with the same set of data in Partitions ?
(4) Do We Now have a successful Kafka running version Without ZOOKEEPER Dependency ?? which Version is it ?? How successful this new version in market is now ??
Please reply with your insights and answers.
Thanks keep doing this nice work and knowledge sharing for a lot of aspirants to become successful IT careers like us. 👍🙏😊
1) broker is nothing but a server, which we start with a command. Now it can stop or crash any time just like our normal servers.
2) using epoch number. The most recently and most updated isr / broker is choosed az as new leader
3) it's a best practice to use at least 2 replica for fault tolerance. You can have 0 or you have 10
4) we are in 3.x version as of now. 4.x.x will have zookeeper completely removed
Great Video
Thanks 👍
Many thanks❤
😊😊👍👍
Awesome ❤, need other topics too
👍👍😊sure
Thanks for video and explaining Kafka very well. very good content.
Thank you very much
Great informative video, Do share the Part - 2
Awesome
Thanks
How to handle the scenario where a consumer gets delayed reading a message, for example- Consumer B reads "booking successful" message before and a Consumer A could read "payment failure" message ? When both messages are from different topics
I am not sure if I understood your questions completely how kafka will manage if data got consumed in different order like order placed is consumed without payment message is received. I will suggest there will be two different microservices one will consume data from payment topic and once confirmed payment is done it will publish data to a different topic from where booking service will book the order
Now second scenario if there is single microservice consuming from a single topic then it's code responsibility to check payment status first then go for booking.
Please correct me if I am wrong
Your presentation is very good. 👍 There are so many videos on Udemy which don’t follow minimum standards. I really liked the way you explained the concept. I am not sure if you have your training courses on udemy as well. If not you can try that and you will be a star teacher over there. Just one suggestion though, don’t repeat a point of explanation multiple times, to make your video crisp and clear.
Thanks man. Means a lot 😊😊
Thanks very useful. We need more interview questions on kafka.
Sure 👍👍
Thanks for this good video.
☺️☺️
@CodeDecode Can you please add part-2 video to learn more
Sure 👍
Please create more interview question...
Sure we will create it soon
Please share more interview questions
sure we will create more soon
Pls make part 2 and also pls explain clearly about the consumer groups in nxt video.
Sure we will create it soon
please make more practical videos on kafka, Thank you so much
Sure we will create it soon
Waiting for this long time, Thanks. First comment
Thanks for your comment Akash 😊👍
Brocker will create automatically? Or do we need to create automatically
Hi Mam, Thanks for an amazing content. Requesting you to create more interview videos on Kafka.
hi, it was a very helpful video. If you could please make more videos to make more questions on kafka, it will be extremely helpful.
Hi Mam, please create a video by covering all Kafka related questions for interview perpose
Sure 👍👍
Please continue interview series.
Sure 😊👍👍
Pls make video on basis of cloud like clusters, pods, etc
Sure ekta we will create videos soon on kubernetes concept
Please upload more and more question
sure we will upload it soon
Please create part 2 video
are you from wipro bangalore ?
Nope Pratik !!
kindly cover the remaining topics.
Sure 👍
How much charge and duration ?
Sorry didn't get you question
Please make video on grafana and kibana and Zipkin as well
where is part 2? please send the link
Hi please clarify that how can u ensure that Kafka consumes the messages in the sequential? 2. Will publisher publish duplicate messages? 3. How to retailn failed messages from the dlq within 7 days?
🙏👍
😊😊
Nice vedio on Kafka, but there is a small mistake. You said both Producer writes and Consumer reads from Leader partition, which is incorrect. As this would stop the concept of parallelism when there are more than one consumer, and both trys to read from leader in your case.
Generally, the leader status of the partition does not have any bearing on the consumer.
The leader status only applies to the producers and the broker, so the consumer is fine to read all its data from both the follower Partition and the leader Partition.
"The leader handles all read and write requests for the partition while the followers passively replicate the leader"
As per the documentation of Kafka itself.
kafka.apache.org/081/documentation.html
To answer your question on parallelism,
"Each server acts as a leader for some of its partitions and a follower for others so load is well balanced within the cluster."
Again as per the documentation 😊😊
@@CodeDecode Thanks.. I could'nt able locate the info provided in the documentation..
But for simplicity, lets say we only have one topic, three brokers and 3 consumers belongs to different consumer group.
Now per the documentation only one Broker can have a leader for this one Topic and rest 2 are followers. (I assume we both agree on this)
In this case, lets say reads happens only from the leader, then all three consumer would read only from the leader for this topic, which stops parallelism right.
So ideally for Kafka to support parallelism here, one broker is assigned per consumer, so one of the consumer may read from leader , but the other 2 has to read from the follower. Otherwise in your case, all consumer will read from leader, which is from same broker, which distrupts parallelism.
so here's a more accurate representation:
"The leader handles all write requests for the partition, actively replicating the data to its followers. While the leader is responsible for coordinating writes, both leaders and followers can handle read requests. Consumers can read data from both the leader and followers, providing fault tolerance and parallelism."
please correct me if my understanding is not correct .. still learning, and thanks for the reply..
@@MohaideenA you are correct , @Code decode got this one incorrect..
Your explanation seems reasonable to me too but am still confused as documentation is contradicting it. Give me some time. Will try to connect to open source team. Either they will revise the documentation or else our doubt will be cleared
@@CodeDecode Thanks for getting back to me..
When is part 2 releasing ?
Order of mesages in multiple partitions
How the order of messages is kept , when we have few partitions ?
👍
owsm
Very nice explanation can you please share the new coupon code for this course Master Spring Boot Microservice & Angular K8s CICD AWS??
Can you please cover the remaining topics of Kafka..
i found lot of redundant sentences in this video . reading script in hurry sorry please improve it from next video
Your pronounciation of "T" alphabet is so disturbing
very helpful mam. pls share contact
Thanks. Ucan connect with us on codedecodebusiness@gmail.com
Thank you for this video. Eagerly awaiting for the Part-2