This guide covers what to include in your support ticket so our team can help you as quickly as possible.
Getting Started
To submit a support request, click Submit a Request from the top of this page or visit support.cockroachlabs.com. You'll need to sign in with your registered account.
When creating a ticket, fill out all available fields including Product, Cluster ID, Request Type, Environment, Priority, Impact, Version, Subject, and Description.
The more detail you provide upfront, the faster we can help — every missing detail can add a reply cycle and delay resolution.
What to Include in Every Ticket
1. A clear summary of the issue
Describe the observed behavior, symptoms, or question. Take us through the story of what happened and what you expected to happen. Consider: if you were receiving this ticket, what would you want to know?
2. Timeline of events (in UTC)
Walk us through what happened and when. Include any changes or actions taken before, during, and after the issue appeared. Format like this:
- 2025-06-13 05:04 UTC — Job X begins
- 2025-06-13 05:07 UTC — CPU usage begins to grow on node Y
- 2025-06-13 05:10 UTC — Query performance begins to degrade
- 2025-06-13 05:15 UTC — Applications are shut off
3. Investigation & remediation steps taken so far
Even if they didn't resolve the issue, tell us what you've already tried. This prevents us from suggesting steps you've already taken.
4. Cluster configuration details (Self-Hosted)
Include relevant hardware and topology information:
- Per-node CPU, memory, and disk specs (e.g., 8 vCPU, 32 GB RAM, 500 GB SSD)
- Deployment method (K8s, bare metal, VMs)
- Multi-region configuration, if applicable
- Any custom zone configurations or cluster settings
5. Specific requests and close criteria
Be specific about what you need from us. What questions need to be answered or what needs to be resolved for this ticket to be closed? Keep tickets focused to one topic — if you have multiple unrelated issues, open separate tickets for each.
6. Links to previous conversations
If you've discussed this issue over Slack, email, or in a meeting with your Customer Success Manager or Enterprise Architect, link or summarize those conversations. Also reference any related previous support tickets.
Data to Upload
Always include: Debug Zip (Self-Hosted Only)
A debug zip is almost always essential. It collects diagnostic data from every active node in your cluster, including logs, system info, and configuration. See: How to generate a debug zip
Use --files-from and --files-until to capture the issue window plus buffer time before and after. For example, if instability occurred from 05:15 UTC to 10:47 UTC, gather data from at least 04:00 UTC to 12:00 UTC.
Example Command:
cockroach debug zip ./debug.zip --certs-dir=${HOME}/.cockroach-certs/ --files-from='2025-06-13 04:00:00' --files-until='2025-06-13 12:00:00'Consider including: Time Series Dump (tsdump) (Self-Hosted Only)
A time series dump collects historical runtime metrics from your cluster, including those exposed in the DB Console Metrics pages as well as internal metrics. This is especially important for:
- Root Cause Analysis (RCA) requests for past incidents
- Tickets where metrics have changed or behavior has shifted over time
- Situations where you need to show trends in CPU, memory, latency, or storage
See: cockroach debug tsdump documentation
Use --from and --to to capture the relevant time window. The dump includes 10-second resolution data (within the last 10 days by default) and 30-minute resolution for older data (up to 90 days).
Example Command:
cockroach debug tsdump --format=raw --certs-dir=${HOME}/.cockroach-certs/ --from='2025-06-13 04:00:00' --to='2025-06-13 12:00:00' > tsdump.gob
gzip tsdump.gobCompress the file with gzip before uploading. Upload the resulting tsdump.gob.gz file to your support ticket.
For query performance issues
Statement diagnostic bundle — This is the single most important piece of data for query performance tickets.
- Run EXPLAIN ANALYZE (DEBUG) on the affected query, or
- Use SQL Activity → click into query → Diagnostics → Activate Diagnostics in the DB Console.
Other Artifacts and What We Need
| Symptom | What to Capture | How |
| High CPU | CPU profile (while CPU is high) | Advanced Debug → Profiling UI |
| High memory | Heap profile (while memory is high) | Advanced Debug → Profiling UI |
| Process appears stuck | Goroutine dump (while stuck) | Advanced Debug → Profiling UI |
| Metrics changed / RCA request | Time series dump (tsdump) | cockroach debug tsdump (see above) |
| Reproducible error | Detailed reproduction steps | Include scripts, schemas, and statistics |
Choosing the Right Request Type
When you submit a ticket, you'll be asked to select a Request Type. The description field will auto-populate with a template to guide you:
| Request Type | When to Use |
| Incident | Active production issue — something is broken or degraded right now |
| Troubleshooting | Investigating a problem — something isn't working as expected |
| Enhancement | Feature request or product improvement suggestion |
| Question / Other | General questions, how-to guidance, or anything that doesn't fit above |
Tips for Faster Resolution
- One topic per ticket. Multi-topic tickets take longer to resolve than separate, focused tickets.
- Upload data early. Don't wait for us to ask — attach your debug zip, tsdump, and any relevant diagnostics when you open the ticket. You can use our secure upload tool (SendSafely) for large files.
- Use UTC timestamps. Our team works globally, so UTC keeps everyone aligned.
- Update the ticket if the situation changes. New symptoms, workarounds found, or severity changes — let us know in the ticket thread.
Comments
0 comments
Article is closed for comments.