Questions for your existing IT service provider: (Get the list here to download and print)
Do you have a current network diagram, depicting
all of our network equipment and how it is connected? 
The network diagram has many functions: First, if the technician is trying to walk an onsite employee through a troubleshooting session, and the diagram exists, both of them can look at it and be working together, instead of trying to figure it out and then solving the problem. Second, it decreases the amount of time required to troubleshoot the network when there is a problem. Third, when the network equipment is clearly identified in a diagram, the possible sources of the problem become apparent and easily isolated for troubleshooting and solutions.
Are the jacks and cables clearly labeled?
When tracing cabling, especially end to end problems, identifying the correct cable and jack and testing it will prevent interruption to any other network devices, when it is just one device that is failing.
Is the cabling properly routed with no stresses and strain?
Cable stress is the primary reason for failure. When cables are stressed or strained, the cables pull out of their connectors, first causing intermittent failures, and eventually causing permanent failure.
Do you have a network document that contains information about our Internet Service Provider, our network addressing, system names, network applications, and usernames and passwords?
During an system emergency, no one wants to be searching everywhere for different pieces of information that are vital to getting the network working again. Having all of those pieces of information consolidated is an important aspect of putting the pieces back together again.
Can we get remote support from you?
Remote desktop support options have become very secure and the preferred method to solve user desktop issues. 90% of all on-site service calls can be solved when the technician has visual access to the user desktop; it decreases the time required to solve user problems, and allows for almost immediate assistance.
Do we have full, up-to-date anti-virus software on all of our servers and workstations?
Viruses, spyware, and malware are the primary causes of workstation and server performance problems. In addition, they have the potential of transmitting your confidential business information to unknown and unwanted entities. Performing frequent scans, having real-time protection, and reviewing the results are valuable activities, when coupled with up to date, well-performing anti-virus software.
Are our applications all documented, with the software keys and the PCs that they're installed on?
Having a collection of application CDs is valuable only when coupled with the correct keys that are needed for installation and registration of the application. Knowing which program is installed on which workstation or server is a good inventory practice, and prevents multiple use of the same software license and the potential conflicts that occur in that situation.
If we had a disaster (server crash, workstation failure), how would you get the system restored?
Clean recent backups of workstations and servers are only valuable when the system that crashes can be restored. The best practice today is to create images of a system that can be restored to either the same type of computer, or a different computer if necessary.
Are our documents safe from intruders?
Protection of systems using anti-virus software is the first line of protection from intruders. Securing the systems from outside users by way of password protection, limiting the number of users with remote system access, and using firewall and intrusion detection devices are other manners of protecting your confidential company data from unknown entities. Limiting the use of USB flash drives, external hard drives, and CD writing are new methods for maximizing the data security. Data encryption is another way to prevent unwanted people from viewing confidential files.
Is our router configuration documented?
The network routers (or gateways, or firewalls) normally have a complex configuration that, if it fails, has to be replaced. If the identical device is available, then a software backup of the configuration can often be restored. If an identical device is not available, then the new device has to be programmed with the same parameters and paths that the failed device had. Any time a change is made to the routers/gateways/firewalls, a backup of the configuration should be made, and the change should be documented on paper.
How long would it take to rebuild our entire system if there were a disaster?
If images of the servers and workstations have been created, and recent data backups are available, servers can be rebuilt onto new hardware in between 2-4 hours, and workstations can be restored in 1 hour or less, and if network diagrams and documents were created, an entire new "shoestring" network can be rebuilt in 12 hours or less. This is a disaster recovery exercise, and one important aspect of this question is having offsite storage for the image and data backups, so that in the case of fire, theft, flood, or earthquake, the data is still retained in an unaffected area.
How are our backups protected?
Secure offsite backup storage is recommended. The offsite storage can be as simple as a trusted employee taking a backup tape or removable disk home, or as complex as having the data stored securely at a data center server in a different city, state, or region.
How often are our backups tested?
Untested backups are the leading cause of data restoral failure. The minimal frequency for restoral testing is quarterly, which tests these important aspects of the backups - 1) the integrity of the data. 2) the condition of the backup media (tape, disk, CD) 3) the process of restoring data. Best practice for testing restoral using fixed procedures is monthly, and can be done with a simple set of backup data.
If we asked you to improve your response time to service requests, how much faster a response would we see?
If the response time that you experience is already acceptable, then any increase would probably also be acceptable. The key to this is that every provider could do better depending on the service level that they are asked to respond to. A service level agreement with your current provider is a good way to separate the business-critical problems from the problems that are tolerable or can be worked around.
What is our server status with regard to disk storage, RAM utilization, and CPU utilization?
You want to know when your servers are running low on storage, or taxed to the point where delays are preventing users from performing their work. Monitoring server resources on a constant basis is preferable, but even occasional snapshots of the server resources is acceptable if your technician knows what the trends and thresholds are, and can identify the point where the servers need to be upgraded or replaced.
Are the configurations for our applications and the interactions between our applications all documented?
All applications have interfaces, whether a user interface or an interface to another program. The knowledge of how those applications interact and how to support that interaction is a vital support document that will be invaluable if there is ever an issue with the application or the server and a rebuild or reinstallation is required.
If you were ever out of commission, would another technician be able to step in with the information that you have assembled?
An important question, because providers have life events just like everyone else. If this answer is unsatisfactory, begin the search for a new technician. Call us at 888-273-1445 to schedule an appointment with us and we'll get your network running to it's most reliable performance.
Checklist for your peace of mind
Network diagram
Network document
ISP telephone number and account number
Usernames and passwords for internal and external company-based accounts
Software CDs and license keys
List of software licensed to each PC and server
Data Backup frequency
Data Restoral testing period
Server status sheet (look for trends in utilization)
Service level agreement or memorandum of understanding