Cloud Infrastructure Management

• Server process & resource monitoring

We monitor Memory, CPU and Disk Usages, in the instances. The system automatically sends alerts when these parameters reach the preset levels. Services in the server (like Apache, Tomcat for Java) are also monitored. Scripts are written which attempt an automated recovery/restart of these services in case they are down. An alert is sent if these scripts do not achieve a recovery within specified time duration. Monitoring & analysis of the performance of application using third party tools like Appdynamics, Newrelic etc are also offered on need basis.

• URL monitoring with alerts & built-in escalation

Each important URL is monitored 24X7 and phone alerts sent in case it goes down or takes more than 30 seconds to respond. The alert algorithms also have an escalation matrix and it automatically escalates to multiple levels (max four levels of escalation), if the issue is not resolved within 15 mins. For each down time Neev’s team will report with root cause analysis and corrective/preventive actions taken.

• Monitoring Dashboard

A monitoring dashboard with resource usage visualizations is provided. The usage stats are collated on regular basis to a central server and a history maintained. This can be used for analyzing the app performance, server resource consumption, uptimes, errors and others. Neev also provides monitoring of specific errors in the server/application logs and also implements custom checks for the application. Custom check metric can be defined by the customer as per business requirement.

All customers get access to the monitoring dashboard which typically includes the information about the following:

• Server status
• URL Status
• Website Health
• Error Stats
• Resource Consumption
• Business Objects (Custom checks)

All customers can login on to an online support ticket system and create tickets with defined priorities. The highest priority will trigger a phone alert to the cloud administrator on-call and such alerts have an escalation matrix bound which automatically escalates if not addressed in 15 mins at each escalation level (max of four escalation level).

As per customer’s business needs backups for relevant data at regular frequency is configured. Multiple-day retention is maintained for the same. The backup cycle is flexible on request. Restoration from backup is done when a request is raised.

For each application the relevant OS/software security updates are configured. The security firewall is also managed to control attacks.

The contact details of a senior cloud admins will be provided. For all emergencies we encourage the customer to co-ordinate with the contact person directly.

Design and implement architectures with High Availability and scalability factors using the resources provided by the infra providers. For instance, in case of AWS this will be:

• ELB to load balance the traffic between multiple app servers
• Setting up Multi-AZ deployments provide enhanced availability and durability for Database.
• Configuring auto-scaling for scale your Amazon EC2 capacity up or down automatically according to conditions.
• S3 storage for highly redundant storage
• CloudFront to distribute content to end users with low latency and high data transfer speeds
• Route53 for highly available and scalable Domain Name System.
• Elasticache to manage, and scale distributed in-memory cache environments in the cloud.
• SQS for fast, reliable, scalable, fully managed queue service to decouple components in cloud.

All environments that need to be switched ‘ON’ or ‘OFF’ on regular basis can be triggered with a simple control panel. The frequency, nature and characteristics of the environment can be pre-defined like:

• Automatic Nightly or weekends – auto switch off and switch on
• Manual control for env like test envs, demo envs, etc

Hudson + Sonar is configured and maintained for automated builds. The system also generates “Automated Build Status” reports at regular intervals. We also give admin-controlled access for developer, QA and release teams.

Recommendations to save cost on infrastructure include:

• Reducing infra-bill after analyzing Capacity utilization
• Reducing infra-bill of long-duration infra by taking advantage of reservation
• Using spot resources for temporary purposes.
• While serving static content, usage of S3 with Cloud-front vis-à-vis a bigger instance of app server
• Using glacier for secure and durable storage for data archiving and backup.
• Setting up SES/SNS for mailing and notification purposes.
We also offer services to help implement the cost saving recommendations from an infrastructure or application perspective.