1) Purpose
To safely verify that a workflow (e.g., Scan) is progressing, without interrupting it, using process inspection and log monitoring.
2) Scope
Applies to SFArchive systems where workflows are executed by:
NxWorkflow(systemd service) which launchesNxWorkflowEngine -Q<workflowID>
3) Prerequisites / Access
- SSH access to the node
- A user with sudo privileges
4) Procedure
Step 1 — Become root (if required)
sudo su Step 2 — Confirm workflow services/processes are running
Check for the workflow service and engines:
ps -ef | grep NxWorkflow
Expected
- A
NxWorkflowprocess (often-Bbackground) - One or more
NxWorkflowEngine -Q<id>processes (e.g.,-Q2,-Q3)
Step 3 — Locate the most recent workflow log
Go to logs directory:
cd /usr/share/SFS/logs List newest files at bottom:
ls -lrt Identify the newest matching file:
NxWorkflow_<workflowId>_<pid>.log
Step 4 — Tail the workflow log live
tail -f NxWorkflow_<workflowId>_<pid>.log Interpretation guide
Scanning <path>= directory traversal progressingscanObject <file> ...= enumerating objects<file> exists in the current build table.= already indexed/known; not re-archived- Long gaps around large media files can be normal (checksum/IO)
Step 5 — Confirm forward progress without restarting
If log output is slow:
- Check CPU use (engine busy is a good sign):
top
- Check log file size changes:
ls -l NxWorkflow_<workflowId>_<pid>.log
Rule
- Do notrestart workflow if:
- log timestamps are advancing occasionally, or
- file size is growing, or
- engine CPU is active
Step 6 — After scan completes, verify Queue behavior
Once the scan is finished, work should appear in Queue (NxUI) only then. If you have CLI-only access, confirm NxQueue is running:
ps -ef | grep NxQueue
(Per the manual, the queue service polls nxn_queue and launches jobs when resources allow.)
5) Escalation / When to investigate deeper
Escalate from “monitor” to “investigate” when:
- The engine process disappears unexpectedly
- The workflow log shows repeated errors
- The log file stops growing and timestamps don’t advance for an extended period and CPU is idle
6) Do-Not-Do (Safety)
- Do not restart
NxWorkflow/NxWorkflowEnginemid-scan unless you accept restarting the scan process. - Do not assume “not in queue” means failure during Scan workflows.