The old days spending long hours building on premise infrastructure is over. There are so many cloud-based tools, with such a good quality and easiness for deployment, that the valuable sysadmin time cannot be spent implementing complex tools. Prices are accessible, and generally, there are different levels of licensing where companies can find their right fit.
Most of these tools can be cataloged as Software as a service (SaaS)
Focus of New Relic is real-time performance monitoring for applications. It covers the most used platforms as Java, Python, PHP, .NET, etc. The insights the tool can provide are excellent, even in some cases it seems to be magical, for example automatically linking different services of your platform.
Not only application stats can be collected, New relic has a series of agents or plugins to monitor from the OS level to the database level. Although it is not an inexpensive tool.
In addition to metrics, the tool can capture app errors. And using thresholds for the performance metrics and error rates is possible to send alarms, what implies a much more intelligent control of the infrastructure.
Now a days receiving alerts by e-mail is not enough. A tool like PagerDuty can manage alarms from different sources, and send them via e-mail, SMS or phone calls. It is a great tool to centralize events, and smartly distribute them, for example based on an on-call schedule.
The most basic method to receive alters is by email, but more sophisticated sources can be used, the APIs provided by Pager Duty enables ntegration with tools like New Relic and services like AWS, among others.
Log analysis is a central task during the troubleshooting of issues, and even to proactively detect problems.
The Unix power tools, like grep, are ever present, but sometimes even more powerful tools are required. Graphical representation of events could be key to identify root-causes and easy and fast searches save a lot of time.
Loggly can aid the sysadmin in these matters. Although there are excellent open-source tools, like logstash, to deal with logs, it has a quite difficult learning curve, and sometimes it could have too much complexity. Loggly is an SaaS tool, where you can have log analysis up and running as fast as in few minutes.
Confluence Atlassian OnDemand
Documenting, at least for me, is an enjoyable task. Even more with neat tools.
The wiki concept is here since several years ago. There are plenty of free wiki systems, and maybe they can be enough for most of the environments. But if you want to have an enterprise-class tool, I think Confluence is the right choice. Even at an affordable price.
Atlassian also is leader with tools like Jira, which have a very good integration with Confluence
Network diagramming, was always a choice between two extremes: on one side a feature-rich and expensive tool (Visio), on the other, rustic tools with a bad usability.
Draw.io is a quite good tool, with the basic features required to make a good looking diagrams. Although It doesn’t have a wide number of stencils as in Visio, nor is extensible with new ones, if you don’t need such a fancy graphics, this tool is more than enough.
Regarding the cloud features, it allows saving your files in Google Drive, Dropbox and other cloud-storages.
In addition… it’s free
Time tracking and logging is always unattractive for the technical expert, it can be an awful task if the tools available are old, hard to use, or too rigid. Not to mention the multitasking nature of any sysadmin role, what makes even harder to associate each task to a given issue, project, or client.
A tool like Toggl makes the task easier and a bit more attractive. In the web version, with a single click you can start tracking time for your daily tasks. The mobile app does not let you make excuses of not having the computer close.
Finaly a simple tool to keep to do lists.