Active-Active vs Active-Passive Cluster to Achieve High Availability in Scaling Systems
HTML-код
- Опубликовано: 28 апр 2020
- In this video I want to talk over the active active active vs active passive cluster failover configuration for high availability. We will also explain the pros & cons of using an active-active cluster vs using an active-passive cluster.
2:25 Fail-over • Fail-over and High-Ava...
🏭 Software Architecture Videos
• Software Architecture
💾 Database Engineering Videos
• Database Engineering
🛰 Network Engineering Videos
• Network Engineering
🏰 Load Balancing and Proxies Videos
• Proxies
🐘 Postgres Videos
• PostgresSQL
🚢Docker
• Docker
🧮 Programming Pattern Videos
• Programming Patterns
🛡 Web Security Videos
• Web Security
🦠 HTTP Videos
• HTTP
🐍 Python Videos
• Python by Example
🔆 Javascript Videos
• Javascript by Example
👾Discord Server / discord
Support me on PayPal
bit.ly/33ENps4
Become a Patreon
/ hnasr
Stay Awesome,
Hussein 
Наука
Wow your explanation of complex network concepts into easy-to-understand videos are amazing! Keep up the good work!
I loved your explanation. One good thing that I like about having Active-Passive for Cloud Workloads, you remediate the con of secondary backup server/service contributing towards your infrastructure price even if it's not running however with cloud for instance Lambda if that's your secondary server in Active-Passive, you will not be charged because you pay for what you use, i.e as AWS Lambda charges for execution and Duration only, if it's not getting triggered with request, you will not incur any cost and hence Active-Passive is really a great approach to maintain multi-region resiliency specifically in AWS Cloud.
You just perfectly explained everything 😊
Very good tutorial, thank you for putting this together and sharing with the audience. Very informative. Awesome job.
Glad it was helpful!
Is there any docs to implement this on my kubernetes???
Hello Mr Hussein , can HAproxy and Keepalived be in the same server with app (node in this case ) ?
Thank you for the easy explanation
Nice video to explain what active active is! Helped in my job caz I hv no technical background
You may want to learn better about how VIP addresses work. It is NOT REALLY through ARP, as the clients never see any change and the magic happens in the IP stack of the master and backup nodes:
The master node will assign the VIP to its network card and put a IANA MAC address on it. Then it advertises itself as master, which the backup node will see and will know the master is still alive.
When the backup node stops seeing the advertisements from the master node, it will then assign the VIP to its network card, with that same IANA MAC address.
So it is NOT REALLY ARP, as the clients NEVER see a change in MAC address, but it is the Ethernet switch the ONE that knows which port has that IANA MAC address and simply forwards the packets there.
pepeshopping thank you so much! so it is at an even lower layer. I would agree the way you illustrated it is much more efficient since the clients VIP ARP will always point to IANA mac, and it is a matter of the switching at layer 2 to forward the packets to the right host which answers to IANA mac.. this way you don’t need to clear the ARP table..
Ill do more research on the topic, I am pretty sure the research I have done was just one implementation of VIP and VRRP. Ill make a video to correct this, if you have more resources on this plz send them my way. Again appreciate you correcting me .. i love this ❤️
Excellent blog for better SE
Great video! Thank you. Is there any difference at all between Active/Active and Fault tolerant systems ? Both can involve ARP/VIPs and both involves stacking up servers for redundancy in general....
in aws what will be VRRP .. is it subnet value?
Sorry I don't know much about reverse proxy what does it mean to terminate TLS and forward the traffic in @ 1:26 ?
I explain it here ruclips.net/video/H0bkLsUe3no/видео.html
Great video
great tuto hussein thank you , can u tell me how can I become a member in the channel to watch all the content ? :D
great
how about little bit of Quorem ?
If my understandig is correct, failover on active-passive may take a fraction of a second .i.e. the moment master node stops advertising, backup will assign itself the VIP and this process can take as little as milliseconds with keepalived. However, in the case of the active-active where the failover is being handled by the DNS server, it becomes dependent on the DNS's TTL value and can take much longer (recommended TTL is 30 seconds) to recover. By that logic, the failover handling becomes a disadvantage of the active-active. Would you agree?
Shafeeq Rabbani loved the comment for the deep analysis. 👍 good work..
However your statement is only true in case of AA if ALL servers backed by the VIP goes down which is highly unlikely if you have too many servers.
But sure if all servers behind the VIP goes down then the DNS will keep forwarding traffic to it until it discovers that its down which could be seconds to even minutes
when i configure my servers using nginx and keepalived my request goes to one server only ant help with that
Note im using active active HA
I learnt more in 7 days here than 4 years of university x/
Hi ..my question is not related to this topic . My question is what is the difference between agile and devops? Can somebody follow both at same time ? Or should I follow only one ? Please reply I am totally confused.
Hey Charan! Good question, Long answer so Check out my youtube story I answered there.. ill also make a video about it
@@hnasr ok thanks
@@hnasr Hi thanks for video. What is the relationship between agile and devops ? Can one organization implements both agile and devops same time ? I have seen many people who claimed as devops engineers but they just doing system engineer / admin stuff and creating ci/cd pipelines. But they never do any code development. Please give me ur opinion on this present devops engineer roles ( who don't code at all ) ..
what if we have different haproxy in the different network ?
Salman Kalista its ok. The haproxy will have two network interfaces one public one internal
@@hnasr what i mean, how the both haproxy heartbeat works in different public network , different location , like DRC, because different network will causing problem for master and backup VIP.
why do you clap in you videos!!