PROGRAM GUIDE, read 10 Nov 2026
Run both platforms for one full remediation cycle, with one named system of record at a time
Why parallel running is worth the cost
A migration changes three things at once: the data, the ranking and the people who see the queue. If you switch all three on one day, you cannot tell whether a change in the queue came from the platform or from your configuration. Running both gives you a baseline.
For a Cisco Vulnerability Management customer, the schedule has room for it. The Last Date of Support in bulletin EOL15812 is 30 Jun 2028, and a contract that runs to that date leaves time for a parallel run if you start in 2027. The bulletin guide explains how to find your own term end.
Fix the length and the exit rule
Choose a length equal to your longest remediation cycle. If critical items are due in 15 days and the rest in 30, 60 or 90, a 90 day run covers every class once. Shorter runs hide the slow classes, which are the ones that drive audit findings.
Write the exit rule before you start. A workable rule: at two consecutive weekly comparisons, the top 50 findings by new-platform score differ from the top 50 by old-platform score only where the cause is documented, and every open exception exists in both.
One system of record at a time
Both platforms will open tickets if both are connected to your ticketing system. That creates duplicate work for engineers, and they will stop trusting either queue. Decide which platform is the system of record in each week. Connect the other to a test project or leave ticket creation off.
Brinqa's documentation describes ticket sync with Jira and ServiceNow and a workflow engine. Whichever platform you choose, read how its tickets are linked to findings before you switch them on, so a parallel run does not leave orphan tickets. The ticket guide covers the mapping.
The weekly comparison
Every Friday export the top 50 findings from each platform with owner, score and due date. Compare them in a single sheet. Mark each difference with one cause: missing source, different deduplication, different score input, different asset match or unknown. Unknown is the only cause that blocks cutover.
Keep a count of findings seen by one platform only. A steady number is a configuration difference. A rising number is a connector problem. Check the API export routes of the old platform too: the Cisco export API exports asset, fix, vulnerability and finding data in json, jsonl or xml, and you can use the same files as the old side of the comparison.
- Name the system of record for each week in writing.
- Turn off duplicate ticket creation on the platform that is not the record.
- Compare the top 50 by owner every Friday.
- End the run on the date you set, even if one cause is still unknown, and record the decision.
Who needs to know
Parallel running affects more people than the security team. Tell the engineering leads which queue is the system of record, in the same message, every week. Tell the person who produces the monthly report which platform the numbers come from. Tell the auditor, if the period overlaps with an audit, that two platforms held data on the same dates and which one is authoritative for each week.
Name one person who owns the comparison. A comparison that belongs to the whole team belongs to nobody, and the Friday export slips in the second week. That person also owns the log of causes, and decides when an unknown cause has been investigated enough.
Finally, plan the shutdown of the old side. Decide who removes API tokens, who cancels the subscription (after reading the term and the last date of support), who archives the final export, and who tells the vendor in writing that the data may be deleted. A parallel run that never formally ends turns into a permanent second platform.
Cost is the practical limit. A second subscription for two or three months is a line item, so put it in the budget at the selection stage and ask each vendor for a pilot or a short term that covers the run. Ask also whether the pilot term can be converted into the first contract term without a gap, since a gap forces you to reload data.
Finally, agree what happens if the new platform fails the exit rule. The honest options are to extend the run, to fix the cause, or to return to the old platform for as long as support lasts. Write the choice before the run begins, so that a bad week does not turn into a debate about whether to continue.
- One owner for the weekly comparison.
- A weekly note to engineering leads naming the system of record.
- A written exit date and exit rule.
- A shutdown checklist for tokens, subscription and archive.
Sources
- Cisco bulletin EOL15812, read 10 Nov 2026
- Brinqa: sync tickets with Jira, read 10 Nov 2026
- Request Data Export, Kenna Platform API, read 10 Nov 2026
Last reviewed 10 Nov 2026.