Today, the printer server in my office "gone" for almost half day. It quite a chaos as finance staff in their peak season working on their audit. In the chaotic situation, comes so called scapegoat, the IT maintenance guy. He come to the office to meet with my superior but ended up solving issue on printer server. I swear to GOD that most people that time surely will curse him along the way for all the unnecessary stress contribute by the "death" of printer server. It wasn't his fault really but the tense is piling up.
For any system be it computer system or manual system, maintenance is really vital in ensuring unnecessary stress building up amongst users of the system. As we in technological advanced era, everything had been computerized from bookkeeping to customer behaviour. Even the most advanced technology in computing enable decision making assistance tool to run complex analysis within seconds. More advanced a system, more resources will be use thus interdependencies between system component crucial.
Concept like distributed computing, distributed database, mirror server, security patches and SQL queries optimization are among methods to improve percentage of system availability. As the rule of thumb goes, no matter how good any method being used, a regular scheduled maintenance needed to ensure the method in place work as expected.
System unavailability clearly contribute to unnecessary stress to users. Often, works that should be completed on time delayed due to system. Pressure from management to deliver results drive most people insane to meet dateline. System unavailability as always happen in company in frequent case involved email server which resulting in delay of information thus push the work that could be solved on the same day to the day after. The stress from error message popping up for every now and then together with time taken for users to realize that system unavailable by far contribute most to irritate users. Lack of immediate communication by maintenance party frustrated users most of the time since users will keep trying for not knowing that the system unavailable.
In time of peak season, system unavailability is in the top list together with electricity unavailability and "missing" air conditioning. In recent observation that had been done, user motivation drop rapidly when system unavailable. All planned activities need to be revised hence force unnecessary interruption to workload in user responsibility.
Unnecessary stress contribute by system unavailability also derived from top management which required milestone to be achieved by staff. In very competitive industry, man hours spent in finishing projects distinguish between a successful project and a fail project. Extra man hours equal to extra liability and cost to management. In some project especially in construction field, longer time to wrap up a project foresee the company need to pay their client compensation for late wrap up. Top management as always won't be happy for any unnecessary additional cost which could be avoided. Accumulating total man hours, we will be shocking on how much cost that company need to bear for every incident of system unavailability occurred. Scenario-based analysis for man hours calculation effect to cost as per example below:
Example:
Situation: Email exchange server unavailable for 1 and half hour.
Variables:
1) Staff affected = 100 staff
2) Time waste = 90 minutes
3) Pay for an executive per month = RM 2,000
4) Total working days per month = 24 days
5) Working hours per day = 8 hours
Calculation:
100 staff x 90 minutes = 9,000 minutes
|
\/
9,000 minutes/60 minutes = 150 hours
|
\/
150 hours/8 hours = 18.75 days
|
\/
18.75 days/24 days x RM 2,000 = RM 1,562.50
Conclusion: Company wasted 18.75 working days which totaled up to RM 1,562.50 monetary lost for 1 and half hour system unavailability.
From example, we could easily see the accumulated man hours could directly incurred lost to company. In intangible view, late submission of contract document or delayed respond in replying customer inquiries affect company image which in long run affect product or service market share offered by company. Through grapevine and user recommendation in internet may derive on how a company dealing with customer and certainly system unavailability issue should never be brought up to customer as an excuse for any delay. It might be the best excuse given but when it had been used too much, customers trust over brand will deteriorate.
Re-active & Pro-active Approach
Reactive approach as the term define as approach once the incident occurred. Reactive approach will be use by any means since computerized system will have unpredictable problem. Reactive approach required fast and temporary fix which enable the system before further investigation could be done. Experience system technicians and engineers are required in hot seat to cater the problem on site. Immediate communication between system administration is required to lesser the irritation in users' mind. Example of re-active approaches are on-site troubleshoot, customer log problem, quarantine and many more.
Pro-active method as the name suggest include proper planning, scheduled maintenance, scheduled downtime and regular check up. Pro-active approach suggest action on preventing system downtime to occur more than targeted percentage. Since it's involved prevention, the method carefully studied on how the system could be improve in the long run for expansion taking into consideration system availability percentage. It's almost an exhaustive exercise as when the business change, the requirement for system would be affected thus a good pro-active approach would consider on any business change within certain period of time. Example of pro-active approaches are SQL query optimization, data mirror, backup power supply, distributed database and many more.
As conclusion, it's not an easy job to be the maintenance guy but someone need to do it anyway. Get us out from constant agonies.
Thursday, April 03, 2008
Subscribe to:
Post Comments (Atom)
3 comments:
ehem.
let the technical man talks, plz.
First, SQL query optimization is not a way to ensure high availability/disaster recovery. SQL query optimization is to ensure the performance of the server/software/application to the peak.
And that is not something so easy to achieve; especially when the management itself doesn't want to spend money.
This is applicable to many more aspects; in-site / remote support, planning, etc.
If the management itself wanted to be stingy itself, so expect the same shitty support la kan?
tak kisah la pro or post. (there's no such term as reactive actually in the technical world. a downtime is always a downtime.)
True, its not easy to be the maintenance guy but if the users itself (mind the it) don't want / don't have the opportunity to learn on how to use it properly, then, who's to blame?
Blame your own management for digging your own grave for your business.
Its not like we (the technical person) doesn't have the solution in store. I can talk about HADR (High Availability and Disaster Recover) all day long if you wanted to. Its simply on the shelf.
But does the management have the "lock"?? ;)
Well, nothing personal. Its just business.
hehehe. please be my guide. well, i see SQL query optimization indirectly affecting how long user will hanging to the resource. if the SQL provide faster output, then the user should use the resources time lesser than they should, kan? hence allowing more user to access to the resources in a frame of time thus enhancing the availability of the system to mass user. that's how i see it la.
no matter how good a system be, downtime will happen. there are no foolproof, non-hackable system. but i think a proper action plan for sudden downtime would provide better confidence in user for the system. what i want to say is better communication on the system status would be appreciated. if it's the global operation, then communication must be harder compare to communicate on internal operation.
hehehe. i believe that all party involved directly in ensuring good system management. mind you, i once in management shoes, once in user shoes (still in user shoes) and once in support shoes. the key to this is communication. to cross the bridge of technical and lay man term require a "translator" where business analyst should play their role. we need to educate user, pressing the knowledge to them. it's possible by any chance.
yes, management couldn't be more asshole than they already are. couldn't agree more on that.
[url=http://www.payloansonline.com]payday advance loans[/url]
This is the best way to get all your health products online like green coffee, african mango, phen375 and others. Visit now
[url=http://www.prlog.org/12053028-provillus-hair-restoration-company-now-offers-regrowth-solution-for-men-by-provillus.html]Provillus Women[/url]
Post a Comment