Sematext has been offering monitoring and log management for a while now, but in the most recent release
we’ve further unified it, making it really simple for people to go through the well-established DevOps path:
Alert Notification ⇒ Metrics Charts ⇒ Log Inspection for details
Moreover, we’ve made it dead-simple to see metrics and logs in the single view and visually correlate between the two. End result? Much faster troubleshooting ⇒ much more efficient root cause analysis ⇒ much shorter MTTR ⇒ happier DevOps & happier users!
We’ve restructured the UI.
Everything you need can be quickly accessed via tabs on the left. Account section is exposed, and Team-related functionality
(invitations to apps or to account, user roles, etc.) is in its own Teams section.
Dashboards are in top-left.
Metrics, Logs, and Events are right there next to each other. AppMap, NetMap, and a top-like view of all your Servers is under
Infrastructure. Jumping to a specific Metrics or Logs App or a Dashboard can be done quickly from the searchable pulldown in top-right.
If you have access to multiple accounts you can switch between them much like you can switch between multiple Google accounts.
UI looks and behaves differently. Indeed,
we rewrote the whoooooole front-end using modern front-end technologies like React.js and Redux.
We’ve switched to Chart.JS for charts, and contributed a few PRs to it along the way. We’ve made it possible
to show/hide chart legends, change their positioning, to easily add/remove individual timeseries from metrics charts, and so on.
Dashboards have been completely rewritten and are much, much faster and lighter than before.
You can easily create a Dashboard that contains charts from multiple different Metrics or Logs apps. You can also add logs themselves
to a Dashboard, or event charts, or components with custom Markdown in them.
For more details about this check out our blog
When you’re troubleshooting and inspecting logs what do you do? You search and filter logs to narrow things down, to find log
messages that will help you understand what, how, why, or when something happened. The key part here is that you search, and when
you search what you typically see are only logs that matched what you searched for (think grep).
Log Context in Sematext Cloud is much like that.
Once you’ve found a log event of interest a single click will expose its context.
It’s like grep -C NUM, but applicable across all your infrastructure, not just one log file at a time.
Need to have your recent logs searchable for your DevOps team(s) to troubleshoot issues, but also need to
archive old logs for compliance reasons? Many organizations need that and Sematext Cloud makes that simple.
You can certainly simply pick a longer data retention option along with your plan and daily log volume. Doing that makes all
your logs searchable for a long time. If, however, you simply need to store your older logs, so that you can retrieve them
later on, but do not need to search them, you can use S3 Archiving option provided by Sematext Cloud.
All you need to provide are Access Key ID, your Secret Access Key, and the name of the S3 bucket where you want your
logs to be archived.
For more details on log event context check out our blog post.
Alert Rules and related Notification Hooks (aka Integrations) sharing option was updated. Now each Alert Rule and each Notification
Hook is tied to account under which it was created. This way any member of your team will still see all Alert Rules created
under the common shared account. Each team member will be able to use and create new Alert Rules and Notification Hooks, which will
also be visible to other members of the team (but editable only by the shared account owner, or users with admin role). Such Alert
Rules and Hooks will not be visible to App guests though, so your team’s work in the common shared account will not be affected by them.
We’ve changed how Alert Rules & Notification Hooks are handled in Sematext Cloud.
They now live at Account level, not App level, which makes their management much simpler, more intuitive, much more suitable
for modern teams, as described in Alert
Rules & Hook Notification Sharing. The new “Account Default” switch makes it easy to quickly change where alert
notifications are sent for all Alert Rules that are set up to use default account notifications. Very handy for agile teams
who may change their alerting channel preferences over time, have lots of Alert Rules, and would go crazy if they had
to modify each of them.
You can now hook up Sematext Cloud Alerts to a number of different Notification Hooks: email,
any custom WebHook, Slack, PagerDuty, HipChat, VictorOps, OpsGenie, BigPanda, Pushover – for Android, iPhone, iPad, and Desktop (Pebble,
Android Wear, and Apple Watch, too!), Nagios – yes, people still use it, Zappier – and from there you can connect Sematext Alerts to a
number of other services. Setting any of these notification hooks takes only seconds. You only need to configure them once and then,
because notification hooks and alerts are set
at account-level, you can reference them when creating alert rules.
We’ve refreshed our backend, or one of them at least. We’ve updated Elasticsearch to 5.x version. Along with it, we’ve
updated the integrated version of Kibana to which you can easily switch from the “Native UI” if you are a Kibana aficionado. From there
you can create all kinds of charts and dashboards. We’ve also made sure Kibana URLs are shareable, so you can share them
with your teammates. If they have access to the Logsene App whose Kibana URL you are sharing they, too, will be able to access
the shared charts, dashboards, etc.
Unlike logs, which tend to be high volume, events are a little more infrequent, a little more special.
Events are things like alerts, or deploys, or builds, or… Ultimately though, it is up to you and your team to choose what
to push into Sematext as an event. In this release we’ve surfaced Events all the way to the top, so they now stand shoulder
to shoulder to metrics and logs.
You can view logs in their own event stream-like view and you can search them of course. Importantly, however, you can
also correlate events with metrics, as shown below, much like you can correlate logs and metrics. Why is
correlating events with metrics valuable? Think about being able to correlate a bad build or not quite fully tested release
deployment with a change in metric values. When it’s sudden and things break right away you don’t necessarily need any
tools to point out the correlation – you’ve got fire you need to put out and the correlation is obvious.
For such things, it is handy to be able to look at metrics and events and see the correlation. Sounds good?
Log in to see for yourself and check it out!
Logsene, our hosted ELK stack now part of Sematext Cloud,
exposes the Elasticsearch API. If you have experience with Elasticsearch though, if you’re aware of its rich ecosystem of tools, you know how
valuable this is. When you send your logs to a Logsene app their structure (their mapping in Elasticsearch parlance) is automatically
detected. Log event fields are created and their types are deduced from the received data. The mapping and the index template are also
created. Under the hood, Logsene maintains a sequence of indices for each Logsene app. Every so often a new index is added and its mapping
is based on the index template for that Logsene app.
Up until now, Sematext Cloud users had to first provide their credit card details under their
account and then manually assign that credit card to each monitoring or logging App they wanted to use with a paid plan. It turns out
this is a bit of a pain when you have a lot of Apps, so we’ve done away with that. From now on, the payment method defined at the
account level is automatically applied to all Apps with a paid plan. This simplifies management, especially in cases when users
have many Apps (think about credit card expiring and having to assign a new card to each of the 57 Apps you have!).
For more details on this check out our blog post.
We’ve built and open-sourced an Android SDK for shipping logs
from mobile apps to Sematext Cloud. This library not only lets you send your Android app logs
for troubleshooting purposes, but it also lets you collect app usage for analytics purposes! In other words, shipping logs and
usage events from an Android app will give you the ability to both troubleshoot issues with the app, as well as let you gain insights
about your app’s usage, adoption, etc. The library sends data in batches to preserve battery (every 60s), or if there are
more than 10 events queued up.
Want to ship logs and other events from your iOS apps? We’ve open-sourced the iOS SDK for
shipping logs from iOS apps to
Sematext Cloud. This library not only lets you send your iOS app logs for
troubleshooting purposes, but it also lets you collect app usage for analytics purposes! In other words, shipping
logs and usage events from an iOS app will give you the ability to both troubleshoot issues with the app, as well
as let you gain insights about your app’s usage, adoption, etc. If you don’t see the events in the dashboard
immediately, note that this library sends data in batches to preserve the battery (every 60s), or if there are
more than 10 messages queued up.