- 15 Non-Certified IT Skills Growing in Demand
- How 19 Tech Titans Target Healthcare
- Twitter Suffering From Growing Pains (and Facebook Comparisons)
- Agile Comes to Data Integration
Page 3 of 7
Java-based Longitude is easy to use, installs quickly, doesn't require agents, has a middle-of-the-bunch price tag and monitors a veritable plethora of devices, computers and applications. Longitude's simplicity belies its sophistication. We found we could use Longitude not only to keep our network healthy, but also monitor SLA compliance and perform capacity planning analyses.
Longitude comprehensively tracks thousands of operational metrics that it uses in its alerts, reports and charts. It monitors a wide variety of operating systems and environments, software and network infrastructure. Longitude can even keep a watchful eye on user and business metrics.
Longitude's ability to monitor VMware is remarkable. We found that we could automatically collect physical and virtual performance metrics for VMs, hosts, resource pools, clusters, datastores and whole data centers. Longitude consolidated VMware-generated alarms for unified alerting and reporting. We monitored the effect of virtual machines on the physical hardware, and we could optionally take corrective action on any of the performance metrics. Longitude doesn't support Microsoft Hyper-V or Citrix XenServer.
Longitude uses WMI to monitor Windows servers and desktops, SSH to monitor Unix machines and SNMP (via SNMP V1 community strings) to monitor network devices.
We could make Longitude's alert thresholds exactly as complex and realistic as we wished. For example, we used what Heroix terms correlated events to tell Longitude to alert us if a combination of different event conditions occurred. We specified two Event Conditions - file server CPU usage exceeds 50% and file server network connection traffic below 10% -- and then tied these conditions to a "runaway file server process error" correlated event. In a test, we deliberately ran a CPU-intensive program on the file server computer while no one was accessing its files. Longitude dutifully warned us that the file server was behaving strangely.
Longitude can send particular correlated event e-mail alerts to people other than network administrators. For instance, suppose the wireless access point in conference room 5 is not always reliable. In a test, we instructed Longitude to advise conference room users to avoid conference room 5 when other rooms' access points were working, but conference room 5's access point was not.
We were able to associate multiple actions with each alert (for example, sending both e-mail and text messages upon the occurrence of an alert), and we could even suppress events, if we wished, based on date, time, computer ID or the occurrence of a different event.
In addition to notifying you via an e-mail note when it detects a problem, Longitude can also send SMS pages and generate SNMP traps. However, perhaps the best notification is the one that doesn't happen - Longitude can initiate corrective actions at your behest.
For example, when it detected a shortage of available file server disk space, we told Longitude to run a batch file program that deleted .TMP files and did other housecleaning chores on the file server. In another test, we told Longitude to run a script that restarted a Windows service (background process) when it detected that the service had stopped running. Longitude fixed problems nearly instantaneously, long before we could have attended to the problems manually.
Longitude's consolidation and filtering of event logs is a terrific time saver. Plowing through multiple server event logs to locate specific important events is not a fun way to spend the better part of an afternoon, but knowing that critical errors have occurred is key to maximizing uptime and availability. Using pre-built, modifiable filters, Longitude collects event log entries from multiple machines and shows you just the ones you need to see.
In a perfectly platform-neutral manner, Longitude displays the filtered, sorted Unix and Windows server log entries together in the same list. The result truly unifies your system management efforts.
Impressively, Longitude automatically kept itself up-to-date (we could turn auto-update off, if we wished), and it also automatically performed maintenance functions on its network monitoring database.
We found Longitude to be the perfect SLA tracker for documenting the uptime and availability of our servers and applications. Besides monitoring the performance of individual servers, applications and devices, Longitude can take a higher-level view of the network via its SLA feature. Longitude rather neatly aggregated a group of our servers (some clustered, some not) to show, for instance, overall uptime for that group because they logically shared a particular workload.
In another test, when one of three related servers suffered downtime, but the two healthy servers continued to ensure application availability to the business community, Longitude on the one hand accurately and correctly noted the server's downtime on its dashboard and in its monitoring reports.
On the other hand, just as accurately and correctly, its SLA feature reported the overall availability of the shared three-server application as "good." When we tested with an SLA specifying that multiple resources (Web server, application server and database server) must all be available at the same time, Longitude unerringly reported SLA violations when one of the resources failed. Longitude's sophisticated SLA analysis understands the difference between individual server or application monitoring and measuring performance against the terms of an SLA.
We found Longitude's browser-based user interface intuitively easy to navigate and understand. Longitude displays at-a-glance real-time dashboards with pinpoint drill-down capabilities. Longitude's Event Monitor groups events by either device or application, and it can display additional information collected from Windows Event Logs, Syslogs, SNMP Traps and SLAs via its intuitive dashboard. It also scales well - when we simulated the monitoring of a large network, we found we could delegate network segments or specific administrator roles to multiple local Longitude administrators.
When it detects administrator access from an iPad, iPhone or Android device, Longitude displays a mobile app user interface. This interface showed us actual performance data when we received an alert notification, and we used the interface's dashboards to view both summary and detailed status information for servers, applications, devices and virtual machines.
The mobile app dashboard's color-coded pie charts gave us a quick, easy-to-understand picture of our network's health, and we used the dashboard to drill down to see specific data for computers, devices or applications we were curious about. On a mobile device, we could also view (and run) reports, see Event Monitor data or use Longitude's Real Time Performance Monitor.
Heroix Longitude runs on a Windows 2003 Server or Windows 2008 Server machine with at least a 2.4GHz P4 or Xeon processor and 2 GB RAM. Be aware that Longitude is written in Java, an interpreted bytecode language, and needs somewhat more horsepower than a native Windows application.