First identify whether the issue involves setup, connection, tools, pipelines, storage, or rental management. Then check node details, credentials, ports, disk space, and system tasks in order. Each path lists what to check, what to do, and what to include in a support ticket.
Node · Time · Command · Exit code
You do not need to read every document from start to finish. Choose the closest task first, then follow the checks on its card. For order status, expiration, and credential details, use the current console view as the source of truth.
Verify the selected model, data center, and rental term, then wait for automatic provisioning and credential issuance. On first access, record node details and verify the command line, graphical interface, and Xcode environment separately.
SSH is suited to command-line work, automation, and tunneling; VNC is suited to tasks requiring the macOS graphical interface. Verify the node and credentials first, then check your local network and port.
Use commands to record the Xcode version, active developer directory, and toolchain status. Save the current output before changing versions so project issues are not confused with environment changes.
xcodebuild -version OutputConfirm that the runner is online, labels match, and the working directory has the required permissions before running a minimal test task. Do not rerun the entire pipeline first, or the failing stage may be harder to identify.
First distinguish project files, dependency caches, build artifacts, and logs by usage. If space keeps growing, choose +1TB SSD, +2TB SSD, or adjust the cleanup strategy based on actual growth.
The order expiration time shown in the console is the operational reference. Check the status again after renewal. When ending a rental, move out code, keys, certificates, logs, and business data in advance.
Check from the outside in. Select the first failed item and the board will show the highest-priority path. If the previous item is unconfirmed, do not skip ahead and change system settings.
First verify the order status, target model, and data center in the console. Once the records match, continue with credentials, ports, disk space, and system tasks.
Record: Node / Time / Symptom / Recent change
Build failures are rarely caused by a single variable. Record the tool versions and execution environment first, then investigate signing, caches, automation logs, and runner conditions to avoid overwriting the evidence with cleanup.
Run the version query and record the active developer directory. If the project requires a different version, confirm compatibility before switching.
xcodebuild -version
xcode-select -p
swift --versionConfirm that the build task can read the required certificates, provisioning profiles, and environment variables. Keep only names, times, and error summaries in tickets and logs; never submit valid keys.
security find-identity -v -p codesigning
env | sortSave the first failed output, then clean only directories related to the target project. Do not start with a full-disk cleanup, which increases the next build time and removes comparison data.
du -sh ~/Library/Developer/Xcode/DerivedData
df -hBreak the logs down across lane, action, and underlying command. Record the first real error, exit code, and working directory instead of capturing only the final summary line.
bundle exec fastlane lane_name --verbose
echo $?Confirm that the self-hosted runner is online, task and runner labels match exactly, and the execution user has the correct permissions for the working, cache, and build directories.
whoami
pwd
ls -ld .
df -hRenewal solves usage duration, storage expansion solves capacity, and Thunderbolt 5 linking supports specific multi-device workflows. These are independent options; record the current state and target result for each before taking action.
View the current expiration time in the console order record, choose a renewal term, and complete checkout. Reopen the order afterward to verify its status and new expiration time; do not use a local calendar reminder instead of the order record.
Use directory usage and growth rate to assess the need first. Clean rebuildable caches before anything else; back up code, signing materials, build artifacts, and business data before deciding whether to add capacity.
For workflows that clearly require multiple devices to work together. Before ordering, record the device count, connection target, data path, and acceptance method; each linked device is priced separately.
| Add-on | Daily | Weekly | Monthly | Quarterly |
|---|---|---|---|---|
| +1TB SSD | $2.9 | $7.9 | $14.6 | $39.7 |
| +2TB SSD | $5.8 | $15.8 | $29.2 | $79.4 |
| Thunderbolt 5 linking (per device) | $1.1 | $2.9 | $5.3 | $14.4 |
Terms help teams align on operational scope. The definitions below focus on real-world use and distinguish product names, connection protocols, and automation roles.
Restoring access does not end the investigation. Check for unusual sessions, rotate potentially exposed keys, remove temporary files, and confirm that data has been moved out and permissions revoked before offboarding.
View reliability and security notesThe more complete the information, the easier it is to avoid repeated checks. Complete the minimum checks in the task guide first, then submit the node, time, commands, and redacted logs.
Confirm whether the issue concerns connection, the toolchain, CI/CD, storage, or rental management. Record each completed step and result, and do not change multiple variables in the same test.
Include the node and data center, time of occurrence, command, exit code, first error, recent changes, and steps already tried. Redact the logs first.
If you have an active order, use a console ticket so the issue is linked to the node and service record. For pre-purchase questions or if you cannot access the console, contact the service team.
Put “unable to connect,” “build blocked,” or “order status issue” in the subject. Include the node, start time, affected task, and redacted evidence in the body. Never send valid credentials through public channels.
Three model tiers are available across five data centers: Singapore, Japan (Tokyo), South Korea (Seoul), Hong Kong, and the US East. Availability is based on the console's real-time status.