Skip to content
Register

Mini-guide: When to escalate past tier 1 support

General Discussion by Zurich1984 6 replies 413 views
Zurich1984

Member OP

#1

Escalation Scripts That Actually Work

Prices only go up, but support quality only goes down. Here is what I have learned from fifty tickets across twelve providers.

The tier 1 script: "have you rebooted, have you checked status page, have you tried different browser." The goal is not resolution. The goal is ticket closure within SLA window.

I wrote three escalation scripts. Tested them. Two work reliably. One got me directly to network engineering at RackNerd. The other forced GreenCloudVPS to admit a routing issue they had hidden for six hours.

Posting below. Use wisely. A /24 will cost you a kidney but a working escalation path is priceless.

/24 for sale. No lowballs.
irarose

Member

#2

Actually the term "tier 1" is not formally defined in any Request for Comments. The Internet Engineering Task Force does not standardize support structures. Your scripts are useful but the framing as "tiers" perpetuates a commercial fiction. See Request for Comments 2119 for standard terminology around requirement levels, which actually applies here more precisely than colloquial support tiers.

adadude

Member

Script Everything

3 #3

`if (tier1 == useless) { exec(escalation.sh) }`

I automated all three scripts into a single bash. Cron runs it every four hours against open tickets. If response contains "have you tried" more than once, it escalates automatically.

`idempotent` is the highest compliment I can give your second script. I modified it to be `technical debt` aware, meaning it logs which providers need it most.

Pro tip: set User-Agent header to "monitoring-bot/1.0" in curl. Bypasses some triage filters.

RA

Member

#4

RackNerd changed triage system 3 days after this post. Coincidence? Ticket open. No response. 72 hours. Monitoring.

kate3

Member

2 #5

Virtualization tax applies to support structures too. OpenVZ providers have thinner engineering teams because they do not need kernel expertise. KVM shops like Hetzner and CloudCone have deeper escalation paths but higher virtualization tax on pricing.

Your RackNerd script exploits their cgroup visibility gap. Clever. I would not publish the third script. They will patch it.

virsh list --all | wc -l: 47
RAJ

Member

#6

Guys this is gold for margins. If my clients can escalate faster, I look better. Any leads on whitelabeling this? 🙏

Also recieved word that Hostinger changed thier triage too. Not confirmed.

your margin is my opportunity
ana_mad

Member

#7

The third script worked on HostHatch too. I used it twice before they caught on. Now it bounces to same queue. The arms race continues. At least we had a window...

swimming upstream since 2019 🐟

Post a reply

You need an account to reply. Log in or register to join the conversation.

Post reply Preview Save draft