Troubleshooting
We have organized this guide so you can quickly find and resolve issues by symptom.
If systemctl is-active plantpulse-edge.service is active, the gateway is running in container mode — for some issues, the scenario matrix (A~E) in the "Failure Response" section of
Container Mode Operations provides faster diagnosis. For a one-time collection of diagnostic info:
sudo bash /opt/kopens/install/bin/doctor.sh
# → /tmp/pe-doctor-`<host>`-`<ts>`.tar.zst (support 전달)
Quick Index
| Symptom | Where to go |
|---|---|
| The gateway screen won't open | Screen won't open |
| Can't log in | Can't log in |
| Connection fails (PLC) | Connection fails |
| Connected but no values are coming in | No values coming in |
| Values come in but look wrong | Values look wrong |
| Values suddenly stop coming in | Values suddenly stopped |
| The screen is slow | Screen is slow |
| The memory/disk gauge is red | Insufficient resources |
| Data transmission to Cloud / SCADA fails | Transmission fails |
| Tag [Write] doesn't work | Cannot write values |
Screen won't open
This is when you enter the gateway IP in a browser, but the page doesn't appear.
Step 1 — Check if the gateway is visible from your PC
From a command prompt:
ping <gateway-ip>
If there's no response, it's likely a network, cable, or VPN issue. Notify the system administrator that "the gateway is not visible."
Step 2 — Try a different browser or user
If the same symptom occurs on another PC, the gateway itself may be powered off or hung. The system administrator will need to reboot the gateway PC.
Step 3 — Check the URL
http://게이트웨이IP/is the standard format.- There can also be cases with a port number appended (e.g.,
http://게이트웨이IP:8080/). - If you don't know your company's standard URL, ask a colleague.
Can't log in
I forgot my password
Try the default password.
- ID:
admin - Password:
admin123!
If the password has already been changed, request a password reset from the system administrator.
I get kicked out right after logging in
This might be a session or cookie issue. Clear your browser's cookies and try again.
Connection fails
This is a situation where the connection icon on the card in the [Connection] screen is displayed in red (slash over Wi-Fi icon, tooltip says "Disconnected").
Step 1 — Can your PC see the PLC?
Ask the system administrator to verify if a ping to the PLC IP is working from the gateway PC.
Step 2 — Is the port correct?
Default ports by PLC type:
| PLC | Port |
|---|---|
| OPC-UA (Kepware) | 49320 |
| Modbus TCP | 502 |
| Mitsubishi (Q/L) | 5000 or 1029 (PLC setting value) |
| Mitsubishi (iQ-R) | 1500 |
| Siemens S7 | 102 |
| LS | 2004 |
| AB EtherNet/IP | 44818 |
The communication module on the PLC side must be enabled.
Step 3 — Security settings on the PLC side
| PLC | Additional settings required |
|---|---|
| Siemens S7 | Enable "PUT/GET communication" in TIA Portal + turn OFF the DB's Optimized mode |
| Mitsubishi | Enable "MC protocol / TCP / Port" in GX Works |
| OPC-UA | Allow user/password or anonymous access |
| LS | Make sure ICMP (ping) is not blocked |
Step 4 — User / Password
For PLCs that require authentication, such as OPC-UA, the ID/password must be correct. Try re-entering them in the PLC's [Edit] form.
No values coming in
This is when the card's connection icon is green (connected), the collection badge is also normal, but only the tag's [Current Value] field is empty.
Step 1 — Check the PLC address notation
This is the most common cause. The notation differs by PLC type.
- Mitsubishi:
D100,M50(no%in front) - Siemens:
%DB1.DBW0(%in front, DB number + DBW/DBD) - LS:
D00309(5-digit number) - OPC-UA:
ns=2;s=경로format
Please double-check the [PLC Address Notation] section on your PLC's page.
Step 2 — Check the data type / Format Assist
If you have a value that is 32-bit or larger but [Format Assist] is empty, the value may be misread as 16-bit and thus not visible.
- 32-bit integer: Format Assist =
DW - 32-bit float: Format Assist =
REAL - 64-bit float: Format Assist =
LREAL
Step 3 — Verify the same address with another tool
Use the monitor/Watch screen in your PLC programming tool (GX Works / TIA Portal / XG5000 / UA Expert) to verify that the same address is actually working.
Values look wrong
Values are coming in, but the numbers are too large, too small, or the sign is reversed.
32-bit data being read as 16-bit
This appears as random values or as if one side is truncated. Specify DW (integer) or REAL (float) in [Format Assist].
Read unsigend values as signed
Example: 65535 appears as -1. Specify UI (16-bit unsigned) or UDW (32-bit unsigned) in [Format Assist].
Byte / word order (endian) is reversed
Certain slaves (common with Modbus slaves) send values in little-endian, which can corrupt 32-bit floats.
- For Modbus, try a data type notation variant like
:UDINT_LSWORD_FIRST - For other PLCs, post-processing with Apply Formula is often not feasible; contact the system administrator
Different units
For example, the PLC sends an integer raw value (e.g., 1638) but the real meaning is 163.8. Use Apply Formula to enter ${VALUE}*0.1 for unit conversion.
Values suddenly stopped
Values were coming in fine, but stopped at some point.
Step 1 — Check the connection status
In the [Connection] menu, what is the connection status for that PLC?
| What you see on the card | Meaning / Next action |
|---|---|
| Connection icon is red (tooltip "Disconnected") | Problem on the PLC side. Reboot the PLC or check cables/network |
| Collection badge is gray "Stopped" | Someone pressed ◼ or it was auto-stopped. Resume with the ▶ button |
| Collection badge is yellow "Abnormal" | Collection is happening but reads are failing. Look at the [Last Error Message] in the card details |
| Icon green + badge normal but no values | The address on the PLC might be deactivated |
Do not look for the text "Connection Failed" on the card — it's not there. The connection status is the Wi-Fi icon on the right, and the "Connected / Disconnected / Unknown" text is a tooltip that only appears when you hover over it. The text "Normal Connection / Connection Failed" is something else — it's in the network panel on the home screen.
Step 2 — Inspect the PLC itself
- Is the PLC in RUN mode? (If it's switched to STOP, values will stop.)
- Have there been any changes to the memory protection settings for that address area?
- If this happened right after a PLC firmware update or reboot, reconnection should happen automatically after a short wait.
Step 3 — Restart the gateway
Use [Settings] → [Restart Application] to bring the gateway itself back up. It will recover within a minute.
Screen is slow
When there are too many PLCs or too many tags, the initial load may be delayed by 1-2 seconds.
- Generally, this is not a serious issue.
- If the CPU gauge on the main screen is red, the gateway PC's resources might be insufficient. Contact the system administrator.
Insufficient resources
This is when the CPU / memory / disk gauges on the main screen are displayed in red.
| Gauge | Action |
|---|---|
| CPU | Check if any PLC is using too-aggressive a collection interval. Increase it to 1 second or longer |
| Memory | [Settings] → [Restart Application] can temporarily resolve this. If it recurs, request a memory upgrade from the system administrator |
| Disk | [Settings] → [Clean Temporary Files]. The time series retention period might be too long |
Transmission fails
Values are being collected fine, but they are not being relayed to Cloud / SCADA / MQTT.
MQTT / Sparkplug not publishing
- Ask the system administrator to confirm that "MQTT is enabled"
- Verify if the MQTT broker (HiveMQ, etc.) is running on the same gateway PC
No response from the Platform API
- If the PlantPulse platform itself is unreachable → contact headquarters / the operations team
- The gateway itself queues data, so a temporary disconnection will auto-recover.
For detailed log keywords or log locations intended for system administrators, see the Advanced — Developer Notes.
Last resort
- [Settings] → [Download Logs] to download all diagnostic logs in one go
- [Settings] → [Restart Application] to restart only the gateway (1 minute)
- [Settings] → [System Reboot] to reboot the PC itself (3-5 minutes)
If the problem is still not resolved, attach the downloaded logs and contact KOPENS or the PlantPulse operations team.
Cannot write values
This is when you press the [Write] icon at the end of a tag row and you see a message like "Write failed" / "driver returned false".
Write support by protocol
Write support varies by PLC type.
| Your PLC | Write Support | Notes |
|---|---|---|
| Mitsubishi (MELSEC) | ✅ Yes | Q/L/iQ-R all supported |
| Siemens (S7) | ✅ Yes | DB/M/I/Q areas |
| OPC-UA (Kepware, etc.) | ✅ Yes | Requires Write permission on the target node |
| Modbus TCP | ✅ Yes | Holding Register / Coil |
| HTTP push | ✅ Yes (cached) | — |
| LS ELECTRIC (XGI/XBC/XBM) | ✅ Yes | Bit/Byte/Word/DWord/Float/Double |
| AB EtherNet/IP | ✅ Yes | Some types may be rejected depending on PLC firmware/patch |
Common messages
| Message | Meaning / Action |
|---|---|
쓰기 실패 (opc_type=...) | Driver returned false. Check the address format, data type (format), and Write permission on the PLC |
tag not found: ... | Wrong tag ID. Refresh and try again |
scheduler not running for opc_id=... | The PLC is in [Stopped] state. Press the [Start] button and retry |
OPC ... 가 PLC 와 연결되어 있지 않습니다 | Write attempted in DISCONNECTED state — verify IP/port/firewall, then restart |
If it still doesn't work
- Notify the system administrator with the [PLC type + Tag ID + full message].
- More detailed cause can be found in the gateway logs.