# Make Your TAC Life Less Miserable Vendor-neutral tactics for turning ticket purgatory into something vaguely tolerable. ### Dual-wield phone & portal Always file a portal ticket first so you can attach logs, screenshots, and that 3 AM panic selfie. The second it’s in the system, pick up the phone and escalate the priority. TAC agents can’t ignore a live voice reminding them the CFO’s VPN is down rn, and they’ll thank you for supplying evidence up-front instead of drip-feeding it later. ### Credential flex Most vendors auto-triage by the cert on record. Open cases from an account tied to a CCNP, NSE 4, PCNSE, whatever the system punts you straight past the script-readers into Tier 2. No badge handy? Borrow a teammate’s or get a partner rep to lodge the ticket; you’re just using the rules against themselves. ### Break radio silence fast Email threads age like warm milk. If a “high” severity ticket sits un-touched for more than a coffee run, call. SLAs start ticking the moment someone hears your voice, and TAC engineers despise breach reports even more than you despise hold music. ### Pre-ticket triage Spend five minutes in the vendor bug tracker or release-notes RSS before nuking support’s inbox. Logging a “known issue, fixed in 9.9.9” ticket makes you look like you debug with crayons, citing BUG-12345 shows you did homework and often unlocks immediate workarounds. ### Ship the config, scrubbed, not redacted into oblivion TAC can’t replicate a phantom route if you deleted the routing table. Mask passwords and PSKs, sure, but leave interfaces, VRFs, and policies intact. Pro-move: create a temporary read-only admin for TAC, commit, export, then delete after the case closes. ### Demand screen-share early Security policy blocks config uploads? Cool, ask for a remote session right after case creation. A 15-minute live look saves a week of “run this command, send output” volleyball and proves you’re as invested in resolution as they are. ### Packets or it didn’t happen When traffic mysteriously dies, attach both a vanilla packet capture and the platform’s flow trace (`diag debug flow`, `monitor traffic`, `debug dataplane packet-diag`). Include filter syntax, timestamps, and the exact “it broke at 14:07” markers so TAC isn’t spelunking blind. ### Pull the full tech bundle Every vendor has a nuclear “collect-everything” command (`show tech-support`, `execute tac report`, etc.). Run it once, gzip it, upload it. It grabs crash logs, environmental stats, and process dumps aka the stuff you’ll get hassled for anyway, so pre-empt the ask. ### Targeted daemon debug If OSPFd is clearly the gremlin, don’t shotgun debug the entire control plane. Flip verbose logging on just that daemon for 30 s, snag output, toggle off. Smaller logs = happier analysts, faster root cause, fewer “can you reproduce while we watch?” requests. ### Stay synchronous until green After TAC pushes a fix or workaround, leave the ticket open and yourself reachable for at least 24 h. Quick feedback loops let them tweak configs before the post-mortem begins and earns reputation points you’ll cash in on the next 2 AM outage.