Lets refresh where we are! We are discussing about architecting a Citrix environment from Architects perspective. There are 2 parts to it:
1) Assessment
2) Design
Assessment is further divided into
· Operation and Support
· Conceptual Architecture
In this blog, we will discuss assessment of Operation and Support. Each topic will be covered in separate blog.
During the assessment of the
operations and support processes in an environment, the architect should gather
data about the monitoring, systems management, testing, and change control
practices, support environment and user issues and the backup and recovery
plans.
Monitoring and Systems Management
To optimize the user experience,
monitoring and management must be systematically addressed in a Citrix
virtualization environment. During an assessment project, architects assess the
current monitoring tools used, and the process for implementing and managing
them. For example, they should find out which administrators are responsible
for the tools, data, and resolving major issues.
Architects should ask if the
following monitoring tools are being utilized for monitoring and systems
management:
- Health Monitoring and Recovery (Citrix XenApp feature)
- Citrix EdgeSight
EdgeSight
is highly recommended for Citrix environments.
- Resource Manager
- Microsoft Performance Monitor
- Microsoft Systems Center Operations Manager (SCOM)
- Third-party tools such as Process Monitor
Ideally,
in a Citrix environment, the following systems are in place:
- A hardware managing solution
- A server monitoring solution
- Health Monitoring and Recovery
- Citrix EdgeSight
Processes for Monitoring Systems
Management
Processes for monitoring systems
management are relevant to any assessment. An architect checks the monitoring
tools implemented to monitor the health of client devices, hardware, networking
and application usage, and to track metrics.
Architects should ask the following
questions:
- If the environment contains XenApp, are Health
Monitoring and Recovery tests utilized?
- What is used to monitor servers?
- Is a network monitoring tool used?
- What transpires when an alert happens?
From
the Architect
A virtualization architect should
have a watchful eye on the user experience and demonstrate a deep level of
understanding of the users and the administrators responsible for monitoring
the environment.
Testing and Change Control
In Citrix virtualization environments,
effective testing and change controls ensure environment stability, while
adding to the flexibility inherent in the solution. Architects review testing
and change control during the assessment to evaluate the infrastructure for
implementing desktop changes. Items assessed may include lab environment
hardware, test processes, change control tools, documentation, approval
processes and general procedures for implementing changes to the environment.
For example, if a company plans to
upgrade the service pack on the desktop operating system, testing and change
control is critical in order to ensure a smooth process. In addition, a
rollback process should be documented in the event that the planned change is
not successful or cannot be completed within the required timeframe.
A best practice is to schedule
maintenance windows or rollout dates where all changes will be planned, tested
and released. Scheduling maintenance windows increases the consistency of the
maintenance schedule and makes planning ahead easier.
Architects should ask the following
questions:
- Is there a testing environment in place to test changes
before going into production? If so, is it isolated?
It is
recommended to have an isolated and designated test environment that is identical
to the production environment but is specifically used for testing.
- How closely does the test environment mirror the
production environment?
- How many servers are allocated for testing?
- Do production servers have the same configuration as
test servers?
- Are virtual machines used in the testing environment?
- Which change process is used? For example, implement,
test, rollback.
- Is there a formal test process? Which types of tests
are regularly conducted, including scalability, integration and functional
unit testing?
- Are there any negative issues, such as disruption of
users, or scheduling difficulty, associated with the current change
process?
- What is the time interval between change request
approval and implementation?
- Are automated methods used to rollout changes to users?
- Is the XenApp configuration log used?
- Are ticketing requests in-line with the help desk?
Support
The support environment is vital in
any enterprise environment, including one involving desktop virtualization.
However, if the customer has not yet implemented virtualization, an architect
may examine the support trouble tickets and determine which issues can be
mitigated by using virtual desktops. Based on the current environment, an
architect should be able to document the potential and limitations of desktop
virtualization.
When assessing the existing support
environment, architects usually meet with the help desk or support manager to
gain an understanding of users' issues and to review consolidated reports that
can reveal trends or areas of concern.
Architects should ask the following
questions:
- What Service Level Agreements or Service Commitments exist?
- What are the existing support tiers? (Example- Level 1:
Help Desk, Level 2: Escalation, Level 3: Engineering)
- Which support contracts exist?
If an
active Citrix Technical Support agreement exists, XenServer incidents
associated with XenApp farm servers are covered.
- What process is used for incident tracking?
- What is the escalation process?
- What, in detail, are the common troubleshooting steps?
- What happens when users call the help desk? Does the
help desk associate help the person or record the incident for follow up
by another employee?
- Does IT maintain data on how long tickets are open
until they are resolved? What is the average time to resolution?
- If a serious problem arises and additional on-site
support is required, how long does it take for someone to fix the problem,
on average?
- What percentage of calls are related to common issues?
What are the largest ticket categories?
- Which tasks are automated to minimize help desk calls?
- Do support personnel have delegated administration
rights to the Delivery Services Console, if applicable?
Delegated
administration rights in the Delivery Services Console is considered a best
practice.
- Where are the call center locations, relative to the
users?
- How many employees make up the help desk staff?
Gathering specific answers to these
questions about the environment provides the architect with an understanding of
the support environment, which will help with making initial recommendations
and with the solution design as well.
Support Recommendations
General recommendations for support
in a Citrix virtualization environment include the following:
- Document clearly defined roles and procedures to ensure
continuity between support tiers.
- Track all issues through an enforced ticketing system
to prevent circumvention and improve the overall support process.
- Determine the appropriate support model and support
structure for the overall desktop and application virtualization effort
after the design effort.
- Limit the number of administrators and ensure that they
are adequately trained. Curtail individual administrator rights.
- If necessary, repurpose and retrain existing staff
during an implementation of the virtualization solution.
- Ensure that all systems engineers and system
administrators are formally trained in the technologies and methodologies
that support the overall desktop and application virtualization
environment.
Citrix and
Microsoft training are specifically recommended to ensure optimal design and
rapid troubleshooting.
- Train help desk staff to identify whether issues are
occurring on the virtual desktop or the local, physical workstation. Help
desk staff will also need to be trained on troubleshooting basic virtual
desktop issues.
Backup
Backup management is another
important area of the environment to assess. Servers and all important data
need to be backed up, as a best practice.
Architects should ask the following
questions:
- What is the current backup procedure?
- How much redundancy exists for Citrix infrastructure
components?
- What backup procedure is in place for SQL server, if
applicable?
- Is there a backup schedule for servers?
- What is the maximum amount of time the environment can
be down?
- Have all single points of failure been mitigated?
- In the event of a hardware failure, is there a quick
way to recover the servers?
- Are the organization's web sites backed-up?
- Does the environment use IBM Tivoli Support Manager, or
a similar tool?
- Is a centralized backup and recovery manager system
being used?
- Is there a backup restoration testing procedure in
place?
Disaster Recovery
Disaster recovery is a critical
aspect of virtualization projects. If a disaster recovery plan is already
implemented in an organization, the architect should assess the current plan
and make recommendations to improve it.
Architects should ask the following
questions:
- Is there a full check-list of critical system
components and ordered steps for recovery and rebuild? During serious
system outage, having documented steps in a quick format ensures that the
system is built according to pre-defined and pre-tested plans.
- Is the plan updated each time systems are upgraded?
- How often is the disaster recovery plan revisited and
tested?
- How long does it take to execute the disaster recovery
plan?
- What percentage of users can be facilitated at a backup
site?
- Which manual steps are required in order to facilitate
a backup site?
- Has off-site storage of the critical software and data
required to rebuild a system been implemented?
A full
backup of the server operating system and Citrix software should be available.
No comments:
Post a Comment