Long Long Int @LongLongInteger

How dumb are people to let these kind of scammers grow on twitter

Saurav Chaudhary @sauravstwt

a month ago

🚀 Want 1000+ Recruiter Actions on Naukri? Comment “naukri” and I’ll share the exact free guide I used to achieve this. Most engineers update their Naukri profile once… and then wonder why no one contacts them. But here’s the truth 👇 Naukri is still India’s #1 hiring engine,

285 12 126 21K 146

But, I also didn't see any post from you Do you lack empathy? It is easy to mock others @tailwiinder

major tom @tailwiinder

2 months ago

So many techbros I follow have outed themselves in the last few days. Now I have a nice list of people to mute/block/unfollow. Their uninformed and unempathetic takes, their refusal to acknowledge the influence they have, their inability to have an open mind and do the bare

41 65 717 28K 64

@cneuralnetwork Bro, even there are tweets where people said they muted/blocked you long ago... all comments were supporting them that you guys just do engagement farming Lets not give a F about what others thing... its our life... we should live how we want... not to impress others

SIRrrrrr If customer is Bhagwan, then you should stop taking money from your Bhagwans

Striver | Building takeUforward @striver_79

3 months ago

Customer bhagwan hai ♥️🙏🏻

12 4 297 21K 8

@CoderUday Reason

Topic 49: Leader and Follower Pattern: ===================== -Problem with quorum is system must ensure that majority of nodes participate in the operation -The solution is allowing only single server (called leader) to be responsible for data replication to coordinate work

Long Long Int @LongLongInteger

3 months ago

Topic 48: Trade Offs - 1 ========== Caching strategies: ---------------------

0 0 4 209 0

@hellonehha The only profession where we have to study till retirement

@thebiryanidev I want to know what Swiggy commented on this

Topic 48: Trade Offs - 1 ========== Caching strategies: ---------------------

Long Long Int @LongLongInteger

3 months ago

Topic 47: Fixing data inconsistency in Distributed system replicas: ======================================== 1. Read repair: --------------- -During reads, it updates any stale replica data -During read, request is sent to multiple replicas, stale data is identified and udpated

0 0 4 247 0

@techhouseowner Sir... give me some investment suggestions? currently doing MFs... small..medium

Topic 47: Fixing data inconsistency in Distributed system replicas: ======================================== 1. Read repair: --------------- -During reads, it updates any stale replica data -During read, request is sent to multiple replicas, stale data is identified and udpated -Reads become slow as if data is stale, it involved writing -Used for short term inconsistencies Hinted handoff: ----------------- -Allow another node to accept writes as a placeholder until the target comes back online -Hence no data is lost, data is restored later -Used for short term inconsistencies Anti-entropy (Merkel Trees): ------------------------------ -Used for long-term inconsistencies -Runs in background -Database creates hashes in the form of a tree, where the leaf node is a row/document/key-value -Two databases exchange these Merkel trees, if the hash differs, that means their subtree has inconsistent data -Once they found the leaf that has inconsistent data, they can update (for example: LWW) -This checking process can be run in parallel

Long Long Int @LongLongInteger

3 months ago

Topic 46: Conflict Resolution Strategies: ====================== -If we are using distributed DBs, multi-leader replication or collaborative editing conflicts occur -We have to decide which update to keep -Conflict resolution strategies are vital to ensure all copies of the data

1 0 4 272 0

Topic 46: Conflict Resolution Strategies: ====================== -If we are using distributed DBs, multi-leader replication or collaborative editing conflicts occur -We have to decide which update to keep -Conflict resolution strategies are vital to ensure all copies of the data remain consistent. -Causality: The cause-and-effect relationship 1. Last-Write-Wins (LWW) Strategy: -------------------------------------- -Latest update wins -Each write is tagged with a timestamp or monotonically increasing counter -If there is a conflict, system choses the most recent timestamped data -Simple to implement but cost is potential data loss -Relies on synchronized physical clocks -Used in: Cache systems, user settings 2. Vector Clocks: ------------------ -Mechanism for tracking causality and event ordering -By exchanging & comparing vector clocks, we can decide an event happened before, after or was concurrent with another event -If 2 updates are concurrent, we will keep multiple sibling values rather than dropping one, and a reconciliation process would merge them -Overhead is storing and transfering vector metadata -Vector length = number of nodes/replicas -Conflict resolution is still done by application or higher-level algos -Used in: Strong eventual consistency, multi-master DBs, collaborative applications 3. Logical Clocks: ------------------ -Assigning artificial time to events -Example is Lamport timestamp -Lamport Clock: Each process has a counter, increments on each event, when a process sends a message. It attaches its current counter value, and receiver sets its clock to the max of its own value and received value -If event A occured before event B, then event A's timestamp is lower than event B. -Problem is, if 2 events are really independent (no causality), a lamport clock will still asign one a smaller number than other a larger number. Though there is no causality relationship between these 2 events -Vector clocks are more advanced than lamport clocks -Used for distributed mutual exclusion, sequencing events -Use case: distributed logging, global snapshots, situation where we only need order events.

Long Long Int @LongLongInteger

3 months ago

Topic 45: Quorum: ====== Minimum number of nodes that must agree or respond for an operation Quorum(N,R,W) N=Replication factor R=Read Quorum W=Write Quorum Strongly Consistent: R+W>N Eventual Consistent: R+W<=N I am not writing much, because it is simple topic

0 0 1 139 0

Topic 45: Quorum: ====== Minimum number of nodes that must agree or respond for an operation Quorum(N,R,W) N=Replication factor R=Read Quorum W=Write Quorum Strongly Consistent: R+W>N Eventual Consistent: R+W<=N I am not writing much, because it is simple topic

Long Long Int @LongLongInteger

3 months ago

Topic 44: Outbox Pattern: =========== -Used in distributed systems to solve dual-write problem (updating DB and sending msg/event at the same time) -Reason is, if one succeed and other fail, it causes inconsistency -Outbox pattern makes msg/event as part of DB transaction -We

1 0 5 160 1

@pulkit_mittal_ Don't be naive.. do something different Which others r not doing... There r already many videos covering those topics... Take consistency problems list u shared... Solve or explain them on utube

@rohit_jsfreaky Would u mind telling how many yoe u have?

Topic 44: Outbox Pattern: =========== -Used in distributed systems to solve dual-write problem (updating DB and sending msg/event at the same time) -Reason is, if one succeed and other fail, it causes inconsistency -Outbox pattern makes msg/event as part of DB transaction -We will have an outbox table, that stores all the events -DB transaction + adding event to outbox table happens in one transaction -Outbox processor reads from outbox table and adds its to message queue, or send directly to other DBs Remember is outbox pattern is not same as CDC, CDC directly listens to db logs and adds that events to message queues. Whereas for outbox pattern, we have to change application code to add data to outbox table. In fact, Outbox pattern + CDC are used together

Long Long Int @LongLongInteger

3 months ago

Topic 43: Checksums, ETags, MD5/SHA: ====================== Small code calculated from a block of digital data. Using checksum you can know whether a data is changed or not Checksum is fingerprint of the data Generating these fingerprints can be done using 2 ways cryptographic

0 0 5 236 1

Topic 43: Checksums, ETags, MD5/SHA: ====================== Small code calculated from a block of digital data. Using checksum you can know whether a data is changed or not Checksum is fingerprint of the data Generating these fingerprints can be done using 2 ways cryptographic hash functions like MD5 or SHA or error-detecting codes like CRC MD5 is no longer used for security-critical applications as it can be used for deliberate collisions CRC: Cyclic redundancy check -Fast to compute, excellent at detecting common accidental errors but limitation is it can collide, so only used in specific cases -Network packets and disk drives use CRC -Used for quick error detection, for TB of data

Long Long Int @LongLongInteger

3 months ago

Topic 42: Content-addressable storage vs Location-Based Addressing ============================= Content-Addressable Storage: --------------------------------- -Data is accessed by a unique identifier from its content -Every file is stored under an address that is generated

0 0 7 182 1

Topic 42: Content-addressable storage vs Location-Based Addressing ============================= Content-Addressable Storage: --------------------------------- -Data is accessed by a unique identifier from its content -Every file is stored under an address that is generated from the content itself using hash function. -Used when immutable data storage is needed -Useful when data integrity, deduplication and distribution are priorities. -Calculating hash for large files means can be slower -If content updates frequently, CAS will treat each version as new data, causing increasing storage Location-Based Addressing: ------------------------------ -Traditional way, we use file-path or address in memory or URL -No hash comparison -When the file is moved we get 404 -Mutable content at fixed address -Can have duplicates

Long Long Int @LongLongInteger

3 months ago

Topic 41: OpenTelemetry (OTel): Opensource observability framework. Standardizes how applications collect and export telemetry data. Three pillars of observability: Metrics, logs and traces

1 0 7 172 2

Topic 41: OpenTelemetry (OTel): Opensource observability framework. Standardizes how applications collect and export telemetry data. Three pillars of observability: Metrics, logs and traces

Long Long Int @LongLongInteger

3 months ago

Topic 40: SLI, SLO, SLA: ========== SLI: What we are measuring -Service Level Indicator -Metric of service performance -Availability, Latency, Error rate, Throughput, etc SLO: How good it should be -specific target for service performance -Internal reliability goal -Key

0 0 4 155 2