High availability is defined as a website or web service that is aimed to
be available with 100 % uptime. The emphasis will be on clustering, load
balancing, distributing the clusters geographically. Before configuring such a system, it is essential to keep in mind on the following factors
|
1. Cost
2. Hardware / Software
3. Back Up
4. Geographical Distribution of servers
5. Type of content that is delivered
6. User Base |
 |
The following is one of the typical high availability configuration;

Firewall: Traffic from the internet is analysed and filtered via CISCO
Pix firewall. The max sessions and thorough put of the firewall is designed
based on the user base and scalability required in the future.
Load Balancers: This is primarily to distribute load among the nodes in
the cluster. Web / Apps servers in particular.
Web/Apps Servers: This is a fully scalable cluster of web and apps
servers. It is designed in such a way that a new node can be added to the
cluster on the fly without any down time.
Gigabit Private Network: The Web / Apps server connects to the database
server via private network. Connectivity with gigabit speed makes the access to
database cluster as if it is on the same machine without any latency.
Database Clusters: No direct connection to the database server is allowed
except from the application layer. Database clusters are high end monster
servers with RAID 10 minimum configuration.
Backup: Fully scalable NAS / Storage Cloud.
Note : The above schema is a typical sample of a high availability
configuration. Certain components can be added or removed based on the
customer's input.
High performance is defined as a website or web service that is aimed to
be accessed at higher speed and stability even when there is an abrupt spike in
user access. The emphasis will be on high power clustering, load balancing, high
throughput backbone connectivity. Before configuring such a system, it is essential to keep in mind on the following factors
|
1. Cost
2. Hardware / Software
3. Back Up
4. Geographical Distribution of servers
5. Type of content that is delivered
6. User Base |
 |
The following is one of the typical high performance configuration;

Firewall: Traffic from the internet is analysed and filtered via CISCO
Pix firewall. The max sessions and thorough put of the firewall is designed
based on the user base and scalability required in the future.
Load Balancers: This is primarily to distribute load among the nodes in
the cluster. Web / Apps servers in particular.
Web/Apps Servers: This is a fully scalable cluster of web and apps
servers. It is designed in such a way that a new node can be added to the
cluster on the fly without any down time.
Gigabit Private Network: The Web / Apps server connects to the database
server via private network. Connectivity with gigabit speed makes the access to
database cluster as if it is on the same machine without any latency.
Database Clusters: No direct connection to the database server is allowed
except from the application layer. Database clusters are high end monster
servers with RAID 10 minimum configuration.
Backup: Fully scalable NAS / Storage Cloud.
Note : The above schema is a typical sample of a high availability
configuration. Certain components can be added or removed based on the
customer's input.
Hybrid configurations are a combination of multiple cloud instances and dedicated servers or clusters.
When a site or an application is required to maintain backend database in a
dedicated & restrictive environment due to regulatory compliances, hybrid
configurations are used. While the application layer can make use of the
robustness of cloud computing, the database layer can be kept in isolation in a
dedicated environment. Before configuring such a system, it is essential to keep
in mind on the following factors
|
1. Regulatory Compliance
2. Hardware / Software
3. Application Design and portability
4. User Base |
 |
The following is one of the typical hybrid configuration;

Firewall: Traffic from the internet is analysed and filtered via CISCO
Pix firewall. The max sessions and thorough put of the firewall is designed
based on the user base and scalability required in the future.
Load Balancers: This is primarily to distribute load among the nodes in
the cluster. Web / Apps servers in particular.
YBOX Cloud Instances: This is a fully elastic cloud which lets you
deploy and scale computing instances on the fly. YBOX Cloud is developed over
Citrix Xen Server and are based on a minimum of 2.0 GHz processor.
Gigabit Private Network: The Web / Apps server connects to the database
server via private network. Connectivity with gigabit speed makes the access to
database cluster as if it is on the same machine without any latency.
Database Clusters: No direct connection to the database server is allowed
except from the application layer. Database clusters are high end monster
servers with RAID 10 minimum configuration.
Backup: Fully scalable NAS / Storage Cloud.
Note : The above schema is a typical sample of a high availability
configuration. Certain components can be added or removed based on the
customer's input.