data dog is an infrastructure monitoring service it empowers DevOps teams to keep tabs on dynamic mas environments data bird-dog realizes it easy to imagine the health of your infrastructure at a high level by locating lotion or service you can also view it on a more granular server by server grade the first step in setting up data bird-dog is using a combination of the data dog agent our API or our consolidations with third-party works and business at the time of this recording we have over 120 consolidations and we’re adding more all the time the data dog agent can run on cloud stages bare-metal servers virtual machines receptacles and more perfect for customers with cloud or hybrid infrastructures we support the agent on all major operating systems deploying and managing the agent in desegregations may be done through automation tools like cook doll and ansible the worker rallies organisation tier metrics like CPU utilization disk i/ o network latency and more at a resolution of 15 seconds exerting other methods we can collect metrics down to one second resolution the agent is resilient to network outages by hold metrics during an stoppage metrics are later submitted and backfilled once the network is back online all communication to data dog is over HTTP utilizing port 443 and traffic is outbound exclusively the data dog agent is entirely open beginning represent it’s easy to dig into the code and catch out how it compiles metrics exactly navigate to the checks d folder to see all the Python writes responsible for collecting metrics from your numerous works out-of-the-box integrations extend the agent to enhance and complement basic monitoring data dog integrates with favourite network servers programing language databases other observing implements code repositories theme queues and more some incorporations like Amazon Web Service expect nothing more to configure than penetrating your chronicle credentials or specifying a secure role the AWS integration calls the vapour watch API and can pull in metrics from a variety of services RDS ec2 ELB sqs and more other integratings use the agent to collect metrics from service and implements racing on the host they’re configured expending a simple yamo file that comes with the agent ship practice metrics to data dog is easy expending a few different methods you can use our Breslau api there’s puppy stats d which is like stats d but with calls to instrument your code directly and you can use libraries for your favorite words which wrap both the API and bird-dog stats D once stated dog starts receiving metrics from your legions it’s time to settled them to work data dog volunteers pre-configured dashboards for each installed integration with the AWS integration useds “il be seeing” an out-of-the-box summary dashboard for each configured busines these dashboards accommodate a immediate synopsi of their entire AWS infrastructure including ec2 specimen RDS multitudes ELB s and more you can modify any pre-configured integration dashboard by cloning it data dog likewise allows users to create custom dashboards to envisage metrics from various service and works there are two types of dashboards and data dog time committees are great for troubleshooting and equivalence and screen boards are excellent general status cards you might display on the Tv observer screen cards are excellent for considering the overall health and status of a service or entire infrastructure you can customize your screen committees with widgets like graphs query costs portraits and more conditionally format widgets to create a visual cue for healthy and undesirable doorsteps you can include template variables to meet switching between label remits easy to examine different subsets of the infrastructure too customers CX OS and other interested gatherings entitled to receive a read-only public URL screen cards are intended to act as a jumping-off point when you recognize something of concern on a screen council move to a time board for triage investigation time boards can be created ahead of time or ad hoc to troubleshoot a particular issue time board graphs are in sync and help with Diagnostics by allowing you to pinpoint what is happening across metrics and assistances at the same time simply lag and put your desired visualization and adopt the metric you want to graph by default data bird-dog aggregates this metric over all hosts reporting it you can choose the method of aggregation here next you can narrow down the scope of the metric using labels tags are crucial to the data dog workflow they allow users to envisage metrics and act at various levels of granularity data bird-dog imports tags from AWS and can be added after the fact as well this graph has a tag scope that is shared by all the web servers to visualize the implementation of its of that group users can choose to group the metric by call as well to see a separate time streak per host environment a Z and more append built-in functions and filters to get precisely the result you need or diagram two metrics side by side to see how they compare here we’re comparing CPU across all entanglement servers from the coming week to last week you can also try more complex queries use various metrics by either clicking the advanced button or by editing the raw JSON time committees make it easy to correlate the changing nature of metrics to discrete happens in the infrastructure like a hotfix by searching for occurrences with the term hotfix you can overlay vertical bars on top of Metro graphs to correlate cause and effect here is a more fleshed out term council with other metric visualizations demo you can also choose from bar graphs heat maps top rosters convert graphs and host delineates zoom into a time frame by pull the cursor if you picture an interesting spike or altered in the metric it’s easy to annotate and working in collaboration with teammates using data dogs at notification organisation allows you to send emails and meanings through our countless chitchat and on-call integrations like slack HipChat page or imperative conqueror ops and more because we can’t be staring at hosts maps and dashboards the working day data bird-dog enables you to be alerted on your critical data via checks observers generate alerts like this one the berth to the event stream and apprise the suitable individuals any metrics enter data dog whether they’re part of a standard integration or custom-built may be visualized and initiation alarms to create a brand-new observe entry choose create monitor from any time series chart which will pre-populate some of the observe domains to save period or steer to monitors then prefer brand-new check now select the metric you have the same aggregation options you checked on diagrams median max min and part then choose the tag remit to be included in this monitor you can set up a simple alert on the aggregate price or a multi notify to have a separate monitor for each member of the group metric observers can provoke on come a doorstep or a percentage change regulate the threshold will give the user of visual cue in the diagram preview then assign the observers a appoint and a description that will be included when the alarm is triggered now you can add notify the relevant characters working email or our collaboration integrations optionally you are eligible to determined a policy to escalate the notify if it’s unresolved within a set period data dog offers different kinds of observers so users can be alerted on the most important aspects of their systems brand-new outlier detection peculiarities make it easier than ever to identify any host or group of multitudes that is behaving abnormally compared to their peers it can be helpful to view the overall trends of your checks to help prioritize work to fix availability and performance questions on the services offered that need it most the observer veer report meets it easy to envisage the noisiest alertings legions and services broken down by week daylight and hour the occasion stream aggregates discrete occasions happening across a user’s infrastructure it acts as a living breathing document for all the changes going on in real-time search for occurrences by tag generator or keyword referred happens will be aggregated together frustrating noise and helping to keep things connected criticism and collaborate consuming the same at notification arrangement to refer content two individuals or units collaboration are crucial to a well-run DevOps team and users can invite as numerous teammates to their data dog account as they’d like you’ve just seen how data bird-dog compiles it easy to monitor complex cloud and hybrid infrastructures with dynamic dashboards and alerting thanks so much for watching goodbye
