|
Resources
News & Events
|
|
Monthly Tips, News and Views from Coyote Creek
|
NOVEMBER 2011
|
Presidents Perspective
Aunt Sally, the LAPD and Gmail
 In 2009, amid assurances that Google’s cloud-based email and apps were safe and secure, the City of Los Angeles signed on to move 30,000 City and LAPD employees’ email and calendars to the cloud. Now, after two years as Google’s much-touted “cloud-computing poster child,” more than half of L.A.’s employees still have not been converted. Needless to say, the City does not want to pay for the 16,000 seats that are still on the old system.
The Issue Is Security
Why haven't these seats been converted? A lack of security. Just like there are standards in business or commercial enterprises for information security and protection of evidence, there are very specific standards (i.e. the U.S. Criminal Justice Information Systems - CJISS - policy requirements) that must be met any time a government agency touches criminal records. Unfortunately, Google has not been able to meet these standards.
The "Excuse" Doesn't Hold Up
So what has Google had to say about this? Their basic answer has been to assert that Cloud computing and infrastructure is still fairly new, and the technology to protect and secure data is still emerging. Which seems to me like a fairly glib explanation for what is a very serious problem with Cloud computing. Sure, I suppose that a person with a residential account may not care if their email to Aunt Sally asking what to bring to Thanksgiving dinner gets mingled with emails from other Gmail users. But the LAPD - and, most likely, your organization - deals with sensitive information that cannot get into the wrong hands.
How Do You Feel?
Whether you're dealing with Google or some other provider, how comfortable are you putting your information and your company's information into the Cloud, without first having a very clear understanding of how that data would be protected? I know my answer to the question and Coyote Creek has created a solution to this problem - Coyote Creek RMM for Exchange. If you're interested to learn more about your Cloud and Exchange email options, please click here to join us on Wednesday November 16th at 10:00.
Mike Faster
President
CIO Dinner
Where does SaaS Fit in your Applications Portfolio?
On November 8th at the Four Seasons Hotel in Palo Alto, Coyote Creek hosted our annual CIO Dinner on the topic of Where does SaaS fit in your Applications Portfolio? We invited Todd Fitzwater of Demand Solutions Group to speak and to write a guest blog for us. Todd's blog can be read in full here.
Todd's assertion that SaaS belong everywhere in a company's applications portfolio provoked a lively discussion among the 40 Silicon Valley CIOs that attended the event who were keen - and very open - in sharing their experiences with SaaS applications. Listening to what they had to say, Coyote Creek's top three takeaways are:
- Most CIOs/companies accept that SaaS either is or will become a significant applications platform within their company but, at least in the larger enterprises, it will not become the dominant platform.
- Many CIOs are developing their evaluation criteria for SaaS apps based on and are conventional criteria (e.g. business value, vendor viability, TCO) as well as new issues such as the risk associated with having sensitive data (e.g. IP, employee data) stored outside of their firewalls.
- Some CIOs expressed concern that if everyone used the same SaaS application, how could anyone gain competitive advantage while others took the view that it was a company's ability to execute not the apps they use that gave them an advantage.
Coyote Creek can help assess your SaaS risk and develop a strategy that mitigates risk while enabling you to take advantage of the time-to-value and operational agility that the SaaS model provides. For more information on how we can help, please contact us.
Technical Editorial
Root Cause Analysis: Putting Out the Fire for Good!
One of the most frustrating things for an IT professional is having to battle the same fires over and over again. Root Cause Analysis (RCA) is an important tool for putting an end to this pattern. The key to effective Root Cause Analysis is that there are steps that must be taken while the problem is happening. If you wait until the problem is solved and then say "hey, remember that P1 incident last week? We should really fill out an RCA," it will be too late.
It always starts the same way, the frantic phone call, or page in the middle of the night, something really important is broken, and needs to be fixed. As soon as you get that call, Coyote Creek's #1 piece of advice is: immediately start taking notes and keep a record of everything that happens - you never know what will be relevant and important so don't ignore anything.
Identify the Problem
Quickly assess all the information available and identify the issue, this should be done in the first section of your notes. You just received an alert from your monitoring tool. A critical mail server has a growing mail queue that is much larger than the accepted level for this server. In this case the problem is that there is too much mail in the queues, and if it isn't resolved quickly, very bad things will happened (at worst: crashing server and lost mail, at best: delayed mail and unhappy people).
Detail the Problem
Start recording the details of the problem, the What, Where, When, and Extent. The What is going to be the mail server, the Where is the logical location of the server in the organization's mail flow, the When is the time the issue was first noticed, and finally the Extent is the how far spread the problem is, and it should be noted if this is a customer impacting incident. Once you have recorded the fundamentals of the problem, start gathering more information.
Gather Evidence
Copy and paste any relevant information (on-screen errors, system log errors, CLI commands and their corresponding output) into your notes. Also, be sure to record any unusual system or network behavior you observe. A monitoring tool or monitoring service is critical in this step.
Develop Theories
Review the evidence, and start developing theories on what is causing the problem. For the queue buildup, maybe something is preventing mail from leaving the server, could be mail is flowing into the server faster than it can be processed, or maybe the server is simply underpowered to handle its duties.
Test Theories
Start putting your theories to the test. You believe that mail is stuck on the server, start checking the mail logs or queue viewer to see if mail is going out at all. You verify mail is leaving the server, time to test the next theory.
You theorize that something maybe causing a massive flood of mail to the server. Checking the mail log, you see that there isn't a single source of mail, just lots of mail from different senders and different recipients. It's not a mail flood.
Checking the status of the server, you see the load is really high. Looking for misbehaving processes, you don't see anything out of the ordinary, just many processes related to mail. This starts to point to an underpowered server.
Short Term Fix
The short term focus is to reduce the mail back log. Shut down unused processes/services, temporarily block new mail from coming in, basically do anything you can to get the mail out. As things return to normal, start enabling services, and stop any mail restrictions you enacted. Continue monitoring to make sure there are not any other unseen issues that contributed to the original problem.
Investigate Root Cause
With the short term fix helping to stop the alarms and reduce your stress level, now is the time to find the cause of this incident. Use your monitoring tool to give you a window into past performance of this server. Is the server normally overburdened? Checking the monitoring tool we find the server always has a high load, and the surge of mail combined with other demands on the server put it over the edge. If this over-taxed server is not fixed, you are guaranteed to repeat this incident, and experience some more sleepless nights in the future.
CPU usage on the server
Long Term Fix
So it looks like the server needs more processor to prevent another mail queue build up. If this is a Virtual Machine (VM) simply plan to allocate more CPU resources (verify the hypervisor server has some to spare). If this is a physical server, plan for an upgrade or replacement of the server itself.
Fill out the RCA
With your detailed notes, the process of writing the RCA is simply transferring the relevant information into the form. Of particular importance detail the corrective actions you took when you tested out your theories, even if they action did not appear to help. In the RCA also formalize your long term fix, and assign tasks that need to be completed to implement the fix, don't forget deadlines!
Getting into the habit of making detailed notes and investing a Root Cause Analysis for each major incident can be difficult, but the benefits of not having to fight the same fire again and again, far outweigh the costs.
William Gore
Systems Engineer
|
RMM Webinar
Are Frustrations with Exchange Driving You to The Cloud?
November 16th, 2011
Audience:
IT Decision Makers, CIO, CFO
Microsoft Exchange is the lifeblood system for many organizations. But, as more and more functionality gets added at both the user and infrastructure level, Exchange is becoming ever more complex, time-consuming and costly to monitor, maintain and manage.
If you're considering outsourcing or moving to the Cloud because of Exchange frustrations or cost, join Coyote Creek for this webinar to hear about your alternatives.
To register, please click here.
Product Spotlight
Coyote Creek Website updated using .Net Technology
We're in the process of updating the Coyote Creek website, including creating a new look and feel. After evaluating our technological options, we decided to build the website using the Microsoft .Net platform. Here's why:
- Easy global changes - With .Net's template system, making global changes to colors, fonts, and navigation bars is a snap.
- Dynamic controls - Now we can set and schedule promotions, announcements and other short-term content so that it will only be displayed for a stated period of time.
- Powerful SEO functionality - Including the ability to change page URLs on the fly.
- Seamless integration with CRM - Request forms filled out by site visitors are fed directly into our CRM system.
Want to see our "new and improved" site? Click here. to take a peek.
Contact Us
Your primary contact with Coyote Creek is your Account Manager but you are welcome to contact our management team at any time. Please contact Terri Carney our Director of Sales, and she will get you talking to the right people at Coyote Creek.
About Coyote Creek
Founded in 1998 by a team of former corporate IT professionals, Coyote Creek understands the issues and complexities that are part of large-scale and high-growth IT environments. Pragmatic and vendor neutral, Coyote Creek brings deep expertise to help customers manage technology and project risk, inspiring confidence and creating peace of mind. For more information, please visit www.coyotecrk.com.
|
|