High-Availability vs Fault-Tolerance vs Disaster Recovery

Ki AWS da Auto Scaling Group (ASG) apne aap ch ek 'Fault Tolerant' service Aah ya 'Highly Available' service?

ASG ek 'Highly Available' (HA) service Aah. Jadon server crash hunda Aah, ASG nava server launch karda Aah jis ch kujh minute lagde ne. Is dauran users nu disruption face karni pendi Aah, isliye eh Fault Tolerant (zero downtime) nahi Aah.

Agar tuhanu ek pacemakers (heart-beat monitor) da cloud backend design karna Aah, taan tussi HA chunoge ya FT?

Is critical life-saving system layi strictly 'Fault Tolerance' (FT) chunna pavega. Kyunki agar server fail hunda Aah taan aapan Auto Scaling de nava server banan da wait nahi kar sakde (disruption is fatal). Twin active servers chahide ne.

HA aur FT vich Cost (paisa) da itna vadda farak kyu hunda Aah?

HA vich tussi standby (extra) server tab hi banande/chalande ho jadon primary fail hunda Aah (paisa bachda Aah). FT vich tuhanu primary de naal-naal usda ek exact judwaa (twin) server hamesha 24/7 full power te chala ke rakhna painda Aah (double billing).

What is Redundancy in IT?

Ek hi system ya data de duplicate components (jive 2 routers, 2 power supplies) lagana taaki ek fail hon te dooja available hove. FT aur HA dono redundancy use karde ne.

What is an Active-Passive Setup?

HA da architecture jithe ek server (Active) kam karda Aah te dooja (Passive) sirf tab on hunda Aah jadon pehla fail hove.

What is an Active-Active Setup?

FT da architecture jithe dono servers ekathe 100% time live traffic serve karde ne aur apna data microsecond ch aapas ch sync karde ne.

HA aur FT nu 'Cricket Match' di analogy naal kivein samjhavange?

HA oh Aah jadon batsman injured hunda Aah, match 5 minute layi rukkda Aah (disruption), te pavilion vicho nava batsman aunda Aah (replacement). FT oh Aah jadon batsman de naal hi ek 'Runner' pitch te khada hunda Aah, agar main batsman nu laggi, taan runner turant bhagna shuru kar dinda Aah bina match roke.

HA ch "component quickly replace Hunda Aah" da CloudFormation/AWS vich asali engine keda hunda Aah?

Isda main engine AWS Auto Scaling Group (ASG) aur Elastic Load Balancer (ELB) hunde ne, jo fail hoye EC2 nu terminate karke turant nava template spin-up kar dinde ne.

"halki disruption ya re-login okay Aah" - HA vich user da login expire kyu ho janda Aah?

Kyunki user di asali state (Session data) us purane crash hoye server di RAM vich save si. Jadon nava server aaya, usdi RAM khali si, isliye user nu naye sire ton authentication karni pendi Aah.

FT ch "failure de through operate krna bina kisi user disruption de" physically kivein possible hunda Aah?

Eh 'State Synchronization' naal hunda Aah. Matlab jo vi data Server 1 di RAM ch Aah, ohi same data mili-seconds ch Server 2 di RAM ch vi copy ho reha hunda Aah. Load balancer turant traffic Server 2 te bhej dinda Aah jis kol already saara user session hunda Aah.

Cloud Architecture ch "Uptime" nu kivein measure kitta janda Aah?

Uptime nu "Nines" ch measure kitta janda Aah (jive 99.9% yani Three Nines, 99.99% yani Four Nines). Jinne zyada nines, utna ghat saalana downtime.

IT systems ch "Disruption" da technical matlab ki hunda Aah?

Koi vi aehja event jo user aur server de vich chal rahe continuous network ya data session nu tod dave (break the connection).

"Fault Tolerance" ch 'Fault' da ki matlab hunda Aah?

Hardware, software, ya network de kisi ek hisse (component) da achanak kam karna band kar dena (crash/breakdown).