Integrate Spike with Blackfire.io
Send Blackfire.io alert emails to Spike so alarms open incidents and recovery emails resolve them.
Blackfire.io can notify you by email when an alert rule changes state. Spike gives each integration its own email address, so Blackfire's alert emails create incidents in Spike and the recovery email resolves them.
Step 1: Create the integration in Spike
From the header, click Add integration and select Blackfire.io.
Give it a name and select the service it should alert.
Click Create integration.
Copy the email address shown on the integration page. It looks like
ced6f1c82327db079e3f@email-hooks.spike.sh.
Step 2: Send the alert rule's notifications to Spike
Sign in to Blackfire and open the Alerting dashboard of the environment you want to monitor.
Click Add Alert to create an alert rule, or edit an existing one, and set its metric, threshold and context.
In the notification step, open the channel selection menu and create a new notification channel of type Email directly from it, with the Spike address from Step 1 as the recipient.
Select that channel for the alert rule and save it. Repeat for every alert rule that should page through Spike.
If you would rather keep Blackfire's emails going to your own mailbox, add your address as the Email channel instead and set up a rule in your mail provider that forwards Blackfire's alert emails to the Spike address. Forwarding keeps the subject and body, which is all Spike reads.
Step 3: Test it
Wait for an alert rule to fire, or forward an earlier Blackfire alert email to your Spike address. An incident should appear within a minute.
Example email
Spike receives each email as the fields below. You do not build these yourself; they are shown so you can check what arrives.
Firing:
Recovery:
How Spike reads the email
to / envelope
Yes
Identifies your integration from the {token}@email-hooks.spike.sh address.
subject
Recommended
Match key for repeats and recovery, and the title fallback.
text or html
Recommended
The first sentence about the fault becomes the incident title. html is used when text is empty.
from
No
Shown in the email preview.
Title: for an alarm or warning, the first content line of the email body, for example 96th percentile response time of Checkout on Production is 1,240 ms, above the 800 ms alarm threshold for 5 minutes. Greetings, links and
Label: valuelines are skipped, and the title is cut at a word boundary to 200 characters. If there is no such line, the subject is used. A recovery email is titled by its subject, for example [Blackfire] Recovered: Checkout p96 response time on Production.Emails with no usable subject: an email whose subject is empty, or holds nothing but the tag and a state word, has no key. Its title is the fixed Blackfire alert with no subject (a recovery: Blackfire alert recovered), and all such emails join one open incident.
Grouping: Spike derives a key from the subject by removing the leading
[Blackfire]tag and the state word right after it (Alarm, Warning, Alert, Triggered, Recovered, Recovery, Resolved or Back to normal). Emails for the same alert rule and environment join the same open incident, including a warning that escalates to an alarm. A state word elsewhere in the subject is part of the rule's name, so the rules Checkout and Checkout alert stay separate.Auto-resolve: the state word at the start of the subject decides. Recovered, Recovery, Resolved or Back to normal marks a recovery email, which resolves the open incident with the same key; Alarm, Warning, Alert or Triggered marks an alert. Only when the subject has no state word does Spike read the first sentence, and then only a phrase such as is back to normal or has recovered. An alert for a rule, endpoint or environment whose name contains recovery or resolved is therefore still an alert. A recovery email with no usable subject has no key, so it does not auto-resolve: it is added to the open incident with no subject, which you resolve by hand.
Last updated

