Teams on GitHub Actions: when CI feels slow, where does most of the time go?
1. Waiting for a runner
2. The job itself
3. Re-runs and flaky jobs
4. Keeping runners alive
Reply with a number. If you ran yesterday's check, share what you found.
Tips before you draw conclusions:
• compare p50 and p95, not averages
• split by runner label and by hour of day
• concurrency groups and environment approvals can look like queue time
• step timings are in each job's steps array
One-liner to print both per job, in the image ↓
CI time is mostly time nobody chose to spend: jobs waiting for a runner, hard-to-predict bills, self-hosted runners someone has to patch.
We're building Runzivo to take that off your team, starting with Linux x64 runners for GitHub Actions.
More: https://t.co/1rUOhhwNdR
@MadeByGoogle I have been dealing with Pixel Support in Canada for over a month with 20+ emails and still no resolution.
Google escalated my case to the highest tier of support and created an Advanced Exchange RMA, but Google’s own replacement checkout fails with “Something went wrong.”
I reported the failure and asked for help completing the exchange another way or for someone to call me. Instead, I am now being pushed back to a Standard Exchange that would leave me without a phone for days.
I am not refusing the replacement. Google’s own checkout system is blocking the Advanced Exchange that was already approved.
Case ID: 7-2193000040881
RMA: 71509560
Can someone from @MadeByGoogle please take ownership of this and get it resolved?
Customer service done right. Quick, kind, and generous. Be like Robert. 👏🏾
Get the yearbook for projects and the people who built them.
https://t.co/cnTfO8tni2
If you’re open to moving away from the serverless model, you could run your Next.js app on ZSoftly Cloud Platform (ZCP) in Canada. VMs start at CA$9/month, with no egress fees, so your costs stay much more predictable as traffic grows. Happy to help you migrate and test it: https://t.co/uAKIUTeNpW
Shoutout to the Proxmox team for Proxmox Datacenter Manager.
Once you are running multiple Proxmox clusters, Ceph, and Proxmox Backup Servers across locations, managing everything separately starts to feel like working in the dark.
The value is simple: one place for visibility into health, capacity, performance, backups, updates, security, and what is happening across the whole environment.
This is exactly the direction infrastructure management needs to go.
@moiiikaaa If you're vibe coding until the CI bill shows up 😅
ZSoftly is planning managed CI runners for GitHub Actions + GitLab CI, running on Canadian ZCP capacity.
Keep your pipelines. Change where the jobs run. We're looking for teams for early access 👇 https://t.co/R6zNB2QFgD
Vibe coding is fun until 14 agents are opening PRs and your CI runners are fighting for their lives 😂
We're planning ZSoftly CI Runners for GitHub Actions + GitLab CI.
Managed capacity. Canadian infrastructure. Keep your existing pipelines.
Early access → https://t.co/2hGqfH3iiL
@ojfsaa For dev teams watching GitHub Actions / GitLab CI runner costs 👀
We're building ZSoftly CI Runners: managed runner capacity so you can keep your existing CI workflows without operating the runner infrastructure yourself.
Early access is open:
https://t.co/R6zNB2QFgD
@Shra_va_ni I personally wouldn’t buy a domain from any registrar that locks me into their nameservers. Everyone on this list lets you change nameservers except Cloudflare Registrar — with Cloudflare, you have to transfer the domain out if you want to use another DNS provider.
A small milestone for @ZSoftlyTI .
Tracxn is now tracking ZSoftly in its global public cloud market data. A month ago we appeared at 560/560. The September update has us at 470/560.
The ranking is secondary. Being independently recognized as part of the public cloud market is what matters.
Still building. 🇨🇦