It would be cool if Chicago signed him to a one day contract to retire a Blackhawk. Or do what the Penguins did for Fleury last year and give him a PTO to play a farewell preseason game. The Toews-Bedard connection just once would feed families
#Blackhawks
@jayzawaski Maybe you've addressed it....
People seem to agree that the hawks prospect pool on D is fine. Why? Dmen take forever to develop, and by the time a 2026 pick is ready for the show, one of the current guys may be gone. We have forwards for days in the system.
The worst on-call shifts I've seen all had the same problem.
It wasn't a massive outage. It was 12 hours of noisy alerts.
Things like CPU spikes that didn't affect users, thresholds set too low, pages that trained the team to stop caring.
Pretty soon you run into what I like to call "alert fatigue".
And it doesn't just burn people out - it makes real incidents harder to catch.
Datadog put together a guide on on-call best practices for SREs that breaks down exactly where most teams go wrong:
→ Overly sensitive thresholds that erode trust and increase burnout
→ Paging on infrastructure metrics that don't correlate with user impact
→ No alert classification - so everything feels like a fire drill
→ No structured escalation - so the same person eats every page
The fix isn't just "fewer alerts". It's smarter alerting:
1. Tie alerts to customer experience: page load times, error rates, transaction drop
2. Use composite monitors that combine multiple signals before firing
3. Classify alerts by severity: only escalate what truly needs immediate action
4. Route medium-priority alerts to case management, not someone's phone at 2 AM
5. Run regular alert audits to kill noisy or outdated monitors
The guide also covers rotation design (6-8 engineers, 12-24 hour shifts), handoff protocols (15-30 min overlap, standardized templates), and how to build a learning culture around on-call with blameless postmortems.
If your on-call process is burning people out, it's not a people problem. It's a design problem.
Check out the full guide here: https://t.co/KmIEIzGAV3
Huge thank you to @datadoghq for sponsoring this post.
The sidecar pattern offloads cross cutting concerns like mTLS, retries, routing, and metrics into a co-located proxy. That is a far better approach than forcing every microservice team to reimplement networking and security concerns in application code.
https://t.co/fvb5RIeDYK
Heartbreaking to hear the news today on the passing of @muzz19. Always had so much respect for Troy as a former @NHLBlackhawks and even better person. RIP my friend.
@BenPopeCST Just opted out of my season tickets. The overall product has become AHL worthy both on and off the ice. Be nice if they'd show a damn replay once in a while instead of more crown people. Constant with that and the theatrics. The game is basically an interruption.