This helps you respond and fix the issue quickly so you can keep your uptime numbers high.Ī network monitoring utility such as Spiceworks can also show you: If something does go wrong with a service, network monitoring tools can automatically send you alerts to keep you in the loop, so you can discover and address an outage as soon as something goes wrong. So if you experienced 30 minutes of downtime in the last 30 days, the equation would be. ( time in a given period – downtime ) / time in a given period The formula to calculate uptime for one month is: Some monitors calculate uptime automatically, but even if they don’t, you can simply go back through your logs to see how much downtime you’ve had over a given time period, such as the last 30 days. Going hand-in-hand with an uptime discussion should be network monitoring software, which can keep an eye on all of your servers, services, and websites so you can tell if your service-level agreement is being upheld, regardless of how many nines of uptime you’ve been promised or what you’re aiming for with internal, on-premises services. No matter what decisions you make about pricing and specific apps, the goal will be to make sure IT can provide an acceptable level of service to everyone with minimal interruption. Server and service monitoring: kind of a big deal The Ponemon Institute has estimated that a minute of data center downtime costs $8,851 in 2016, up from $7,908 in 2013. On average, according to a CloudEndure study, downtime costs at least $10,000 a day for 73 percent of organizations. But if the server in question just hosts a website with cute pictures of cats, then some degree of downtime is acceptable.
For instance: if your company were to lose one million dollars in revenue for every second of downtime you experience, or if thousands of highly paid employees would just sit and twiddle their thumbs if a service went offline, then you’d have a business case for higher availability. As the nines go up, your need for redundancy also goes up… and so does your cost.ġ00 percent uptime isn’t realistic, but you can get pretty close.Įxtremely high availability might bring you into cost-prohibitive territory, so determining how much uptime you ultimately need will depend on the application or service in question as well as your company’s budget. Ensuring such a minimal level of downtime can cost you, though. “Six-nines” (99.9999 percent) brings it down to just a few seconds each year. Ninety-nine percent uptime is sometimes called “two-nines,” and when someone says five-nines” (99.999 percent), downtime goes down to a mere 5 minutes per year, or 26 seconds per month. But that might be more than you can tolerate for the mission-critical app running on it.
#BASH INTERNET UPTIME OFFLINE#
That means you could expect roughly up to 3.65 days of downtime, or about 7.2 hours offline each month on that particular server. Say a cloud service provider (Amazon, Microsoft, etc.) guaranteed you 99% availability/uptime over the course of a year for one of your IaaS cloud servers. In other words, it’s the opposite of downtime… the bane of an IT department’s existence. In an ideal world, servers would never fail and we’d have 100% uptime, which is simply the amount of time a service or server is online over a given time period. That is… until a key Exchange server goes down and no one can send or receive email. Most people take fully functioning servers and the applications they host for granted.
#BASH INTERNET UPTIME HOW TO#
Network monitoring tools: How to calculate server uptime