Super explanation sir. I have queries?? You mention that before 11g vips are used for node connectivities. When SCAN got introduced from 11gr2 it resolves connections to the nodes. If SCAN is balancing node connectivity why do vip's divert node connectivities.
Hi Mallik, Good one, If we add or remove nodes from the cluster, how the SCAN gets updated about these details. How to find the ips of SCAN in DNS/GNS.
Hi Sir, we have different nodes(in the sense instances) if request goes from node1 it fetches the data from database and where it will place the data,(in single instance it place in database buffer cache) in RAC it will place all the node's DBC or anything else. Q.one instance goes down,DBC also goes down in this case no data present in that DBC.?
@Gurpreet, VIP has many advantages. 1. Your local listener should run on VIP 2. You application should connect to database via SCAN and then node VIP. 3. Once node goes down your VIP will shift to other surviving node.
2.Even after node addition or eviction on cluster from 11gr2 scan is balancing node connectivities. Then why again vip's are doing user connectivities to surviving node?? can you please let me know sir?
Hello sir, In this video at 9 min 30 seconds , you are saying the "VIP will be shifted to surviving node and the connection is also shifted to surviving node and the connection is still established to database. I am sorry but please correct this statement, it is not true , The VIP will shift to surviving node but the existing connection will not shift to any of the existing nodes. the existing connection will disconnect, user has to establish new connection
Please refer the video on FAN and TAF. I have made one video on that where I have mentioned that you need to setup TAF and FAN so that all existing user or application connection will be shifted to surviving node. Our of those user connected session only select statement will continue to execute but any DDL statement user has to resubmit the request. If you still have doubt in understanding. feel free to call me I will explain you.
@@VismoTechnologies But let's say, We are connected to database using vip of node 1, and if that node goes down, then the vip will failover, but the user will get the timeout error .
i have one question ..if we are adding more nodes into the cluster ..still do we need manually entry in the scan list for those nodes? or will it take automatically and put into the scan list
Oracle Net Services is aware of new addition of nodes. When request reaches it, it will pick one of the db instances on a node to serve data. So, it is implictly taken care while adding/removing the nodes from the cluster.
great session only 1 query. what's the point of having 3 scan ip if it is going to resolve connection in round robin fashion which can be achieved with 1 scan ip as well ?
Guess different subnets for different type of ips can be possible as it will anyway going to resolve the ips to the target system ip. Pl let me knw if this is correct.
If anyone watches all ur RAC videos they can very easily clear interviews of 4 to 6 exp
Thanks Mallik for making best and easy videos of RAC
Thanks a lot Sir, after a long time I got this useful video related to SCAN. You have cleared my all SCAN related doubts in RAC.
You are most welcome
Thanks, this is very detailed and easy to follow
Awesome explaination
Glad you liked it
Thank you for the clear explanation and easy way to simplify the complexity things in Oracle RAC structure.
Waiting for more RUclips videos.
Glad it helped you
Nice One
Thanks 🔥
Thanks very much...this is very useful vedio.
Thank you.
Super good resources to watch!
Thanks Mallik good explanation
Thank you @Naveen, Keep watching my channel and share with you friends.
Super explanation sir. I have queries??
You mention that before 11g vips are used for node connectivities. When SCAN got introduced from 11gr2 it resolves connections to the nodes. If SCAN is balancing node connectivity why do vip's divert node connectivities.
SCAN is for load balancing and HA but user session and application connection has to happen via your node VIPs not with SCAN IPs.
Hi Mallik,
Well explained. where scan ips are mentioned.
Thanks alot..really good job
Thank you. It's greatttttttttt.
Hi Mallik,
Good one,
If we add or remove nodes from the cluster, how the SCAN gets updated about these details.
How to find the ips of SCAN in DNS/GNS.
Nice explanation
Thanks for liking
Hi Sir,
we have different nodes(in the sense instances) if request goes from node1 it fetches the data from database and where it will place the data,(in single instance it place in database buffer cache) in RAC it will place all the node's DBC or anything else.
Q.one instance goes down,DBC also goes down in this case no data present in that DBC.?
Nice Explanation Mallik.
Thanks for liking. Please do watch my videos and share with your Oracle circles
thanks you a lot
Thanks, Malik for a good explanation. What is the role of VIPs after the SCAN introduced in 11gR2.
@Gurpreet, VIP has many advantages.
1. Your local listener should run on VIP
2. You application should connect to database via SCAN and then node VIP.
3. Once node goes down your VIP will shift to other surviving node.
@@VismoTechnologies Thanks Mallik for clearing my doubt ☺
Most Welcome
2.Even after node addition or eviction on cluster from 11gr2 scan is balancing node connectivities. Then why again vip's are doing user connectivities to surviving node?? can you please let me know sir?
@@VismoTechnologies you're awesome man..really helps the explanation..👍🙏
Hello sir, In this video at 9 min 30 seconds , you are saying the "VIP will be shifted to surviving node and the connection is also shifted to surviving node and the connection is still established to database.
I am sorry but please correct this statement, it is not true , The VIP will shift to surviving node but the existing connection will not shift to any of the existing nodes. the existing connection will disconnect, user has to establish new connection
Please refer the video on FAN and TAF. I have made one video on that where I have mentioned that you need to setup TAF and FAN so that all existing user or application connection will be shifted to surviving node.
Our of those user connected session only select statement will continue to execute but any DDL statement user has to resubmit the request.
If you still have doubt in understanding. feel free to call me I will explain you.
@@VismoTechnologies But let's say, We are connected to database using vip of node 1, and if that node goes down, then the vip will failover, but the user will get the timeout error .
i have one question ..if we are adding more nodes into the cluster ..still do we need manually entry in the scan list for those nodes? or will it take automatically and put into the scan list
Oracle Net Services is aware of new addition of nodes. When request reaches it, it will pick one of the db instances on a node to serve data. So, it is implictly taken care while adding/removing the nodes from the cluster.
great session only 1 query.
what's the point of having 3 scan ip if it is going to resolve connection in round robin fashion which can be achieved with 1 scan ip as well ?
It was the easiest steps taught by you sir to keep concept permanent in mind.
Nicely explained 👍👍
Thanks for liking
We can not have different subnet in case of public ips because the different subnets will not communicate each other and failover will not take place
Yeah thats true
sir please can you explain about split brain in oracle rac?
Different subnet is possible but same class of IP (public,virtual and scan).if i am not wrong please correct me...
Yes you are right.
How to configure SCAN ? Any video for that
Answer to your question is VIP and Scan ip should be in same subnet as your public ip
can ip number greater than 255 ? ......... in scan example i think u mistakenly expalined 192.168.1.300 ?
It was just example i gave.
IP address can not go more than 255 in case of IPv4.
If it is IPv6 then you can have IP more than 255 in one subnet
All IPS should be under same sub net mask
Guess different subnets for different type of ips can be possible as it will anyway going to resolve the ips to the target system ip. Pl let me knw if this is correct.
That's true, we can have different subnets.