Operating Cumulocity IoT Edge

Architecture

Networking overlay

The following image shows the networking overlay of Edge server. Two network adapters are attached to the VM with different functionality, as described in the image.

Component overlay

The following image covers the components Cumulocity IoT Edge is built with.

Troubleshooting the system

In case of any issues we provide a number of options which help you find out what might be the cause and give advice how to fix it.

We recommend to follow these steps:

• Perform a health check, see the Health check section.
• Monitor the system processes, see the Monitoring section.
• Check the log files, see the Log files section.

If you still need to contact SAG support, include the output of the diagnostics script. See the Diagostics section on details how to run it.

Health check

Network

Without working network connection the system is not able to work. The following instructions show how to check the network connectivity of the platform.

Check network interface of the node

The following commands will show the interface and network settings of the machine:

$ ip a

This will list all interfaces and their current configuration.

Example:

$ ip a

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host 
   valid_lft forever preferred_lft forever
   
2: enp0s3: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:88:e7:de brd ff:ff:ff:ff:ff:ff
inet 10.0.2.15/24 brd 10.0.2.255 scope global noprefixroute dynamic enp0s3
   valid_lft 85338sec preferred_lft 85338sec
inet6 fe80::a00:27ff:fe88:e7de/64 scope link noprefixroute 
   valid_lft forever preferred_lft forever
   
3: enp0s8: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether 08:00:27:81:fe:9d brd ff:ff:ff:ff:ff:ff
inet 192.168.56.120/24 brd 192.168.56.255 scope global noprefixroute enp0s8
   valid_lft forever preferred_lft forever
inet6 fe80::5b3a:bc65:40b5:f9ea/64 scope link noprefixroute 
   valid_lft forever preferred_lft forever

You need to make sure that the node has an external interface (ethX) and the loopback interface configured (lo). The loopback interface needs to have the fixed IP 127.0.0.1 with subnet mask 255.0.0.0 and the IP address of the external interface must reside in the correct subnet with the correct subnet mask (in this examples 255.255.252.0).

The following command lists the local routing information.

$ netstat -rn

Example:

$ netstat -rn
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
0.0.0.0         10.0.2.2        0.0.0.0         UG        0 0          0 enp0s3
10.0.2.0        0.0.0.0         255.255.255.0   U         0 0          0 enp0s3
192.168.56.0    0.0.0.0         255.255.255.0   U         0 0          0 enp0s8

Make sure you have the destination 0.0.0.0 in the list which then also has the gateway flag (G) set.

Check access to the internet

Try to reach a well-known address in the internet with the following command:

$ ping -s 1500 8.8.8.8
PING 8.8.8.8 (8.8.8.8) 1500(1528) bytes of data.
64 bytes from 8.8.8.8: icmp_seq=1 ttl=56 time=2.61 ms
64 bytes from 8.8.8.8: icmp_seq=2 ttl=56 time=2.80 ms
64 bytes from 8.8.8.8: icmp_seq=3 ttl=56 time=2.82 ms
64 bytes from 8.8.8.8: icmp_seq=4 ttl=56 time=2.75 ms
64 bytes from 8.8.8.8: icmp_seq=5 ttl=56 time=2.79 ms

As when checking the internal reachability you need to make sure that you can see replies from the address you tried to reach. Use Ctrl-C to end the ping command.

Processing

This section lists the required services and processes on the Edge server.

Check platform status

You can check the status of the platform by running the following command:

curl -v http://localhost:8181/tenant/health

* About to connect() to localhost port 8181 (#0)
*   Trying ::1...
* Connected to localhost (::1) port 8181 (#0)
> GET /tenant/health HTTP/1.1
> User-Agent: curl/7.29.0
> Host: localhost:8181
> Accept: */*
>
< HTTP/1.1 200 OK
< Content-Type: application/json
< Date: Sat, 05 May 2018 18:13:28 GMT
< Transfer-Encoding: chunked
< Server: Jetty(8.1.19.v20160209)
<
* Connection #0 to host localhost left intact 
{
    "status": "UP",
    "services": {
        "details": {},
        "status": "UP"
    },
    "mongodb": {
        "details": {},
        "status": "UP"
    },
    "tenant": {
        "details": {},
        "status": "UP"
    }
}

The ‘status:”UP”’ shows that all services are running.

If something went wrong, the endpoint should respond a different error code then 200 OK. The response should contain something like:

{
   "status":"DOWN",
   "services":{
      "details":{

      },
      "status":"UP"
   },
   "mongodb":{
      "details":{

      },
      "status":"UP"
   },
   "tenant":{
      "details":{
         "notFullyInitializedTenants":{
            "savenindia":"PAYPAL_REFUND",
            "nowa":"PAYPAL_REFUND",
            "ttc":"PAYPAL_REFUND",
            "controllogichh":"PAYPAL_REFUND",
            "jonathan":"PAYPAL_REFUND",
            "barcotrail":"PAYPAL_REFUND",
            "ngservices":"PAYPAL_REFUND",
            "sixsenses":"PAYPAL_REFUND",
            "trav2":"PAYPAL_REFUND",
            "lucaslocatrix":"PAYPAL_REFUND"
         }
      },
      "status":"DOWN"
   }
}

The response contains checks for the most important components:

• services - status of OSGI services running within karaf container
• mongo - status of connection to mongo database
• tenant - tenant initialization status

The list describes the tenants which on the core node have not been fully initialized and the initialization status they are in.

Checking the system services

Checking REST API availability

Run the following command to check the REST API availability:

$ curl -u 'edge/<username>:<password>' -v -X GET http://<base_url>/platform
 
* About to connect() to <base_url> port 80 (#0)
*   Trying 52.29.189.245... connected
* Connected to <base_url> (52.29.189.245) port 80 (#0)
* Server auth using Basic with user 'management/<username>'
> GET /platform HTTP/1.1
> Authorization: Basic bWFuYWdlbWVudC90c3NjaHVlbDohITQ3TmV1bjI3MQ==
> User-Agent: curl/7.19.7 (x86_64-redhat-linux-gnu) libcurl/7.19.7 NSS/3.21 Basic ECC zlib/1.2.3 libidn/1.18 libssh2/1.4.2
> Host: <base_url>
> Accept: */*
> 
< HTTP/1.1 200 OK
< Server: nginx
< Date: Tue, 23 Aug 2016 15:39:29 GMT
< Content-Type: application/vnd.com.nsn.cumulocity.platformApi+json; charset=UTF-8; ver=0.9
< Transfer-Encoding: chunked
< Connection: keep-alive
< 
{"alarm":{"alarms":{"alarms":null,"self":"http://management.<base_url>/alarm/alarms"},"alarmsForSource":"http://management.<base_url>/alarm/alarms?source={source}","alarmsForSourceAndStatus":"http://management.<base_url>/alarm/alarms?source={source}&status={status}","alarmsForSourceAndStatusAndTime":"http://management.<base_url>/alarm/alarms?source={source}&status={status}&dateFrom={dateFrom}&dateTo={dateTo}","alarmsForSourceAndTime":"http://management.<base_url>/alarm/alarms?source={source}&dateFrom={dateFrom}&dateTo={dateTo}","alarmsForStatus":"http://management.<base_url>/alarm/alarms?status={status}","alarmsForStatusAndTime":"http://management.<base_url>/alarm/alarms?status={status}&dateFrom={dateFrom}&dateTo={dateTo}","alarmsForTime":"http://management.<base_url>/alarm/alarms?dateFrom={dateFrom}&dateTo={dateTo}","self":"http://management.<base_url>/alarm"},…}

This example shows the correct response of the platform. The username and password need to have full read access to the management tenant. The <base_url> needs to be given to connect to the correct platform and the management tenant must not be blocked from outside.

Monitoring

Monit is a small open-source utility for managing and monitoring Unix systems. It conducts automatic maintenance and repair and can execute meaningful causal actions in error situations.

In Edge, Monit is used to monitor processes and take a restart action if any of them is down/inactive. The usual Monit interval (cycle) is 30 seconds.

Using Monit

Monit can provide a quick status report of all configured services and processes by running the following command as admin user:

$ sudo monit summary

There might be cases where Monit has stopped monitoring some resources because of timeout on constant failures or dependency issues.

A specific component, for example, apama-ctrl_proc, can be restarted using the following command:

$ sudo monit restart apama-ctrl_proc

The Monit status can be checked by running:

$ sudo systemctl status monit

Monit can be restarted by running:

$ sudo systemctl restart monit

The log file for monit is located in /var/log/monit.log.

Log files

Log level locations

The solution stores log files at the following locations for the different nodes.

Core node log file locations

Apama log file locations

To access the apama-ctrl log files, run the command:

$ sudo docker logs apama-ctrl-edge

MongoDB log file locations

Agent log file locations

Adjust log level

This section describes how to change the log level for Cumulocity-specific applications on the backend side. It does not explain how to change log settings for standard components like databases or other operating system related services.

Changing log level for Karaf

The log level for Karaf is defined in the following file.

/usr/share/cumulocity-core-karaf/etc/org.ops4j.pax.logging.cfg

The file has the following structure:

# Root logger
log4j.rootLogger=INFO,out,osgi:*
log4j.throwableRenderer=org.apache.log4j.OsgiThrowableRenderer

# Error appender
log4j.appender.out=org.apache.log4j.rolling.RollingFileAppender
log4j.appender.out.rollingPolicy=org.apache.log4j.rolling.FixedWindowRollingPolicy
log4j.appender.out.rollingPolicy.maxIndex=10
log4j.appender.out.triggeringPolicy=org.apache.log4j.rolling.SizeBasedTriggeringPolicy
log4j.appender.out.triggeringPolicy.MaxFileSize=104857600
log4j.appender.out.rollingPolicy.FileNamePattern=${karaf.data}/log/error-%i.log.gz
log4j.appender.out.rollingPolicy.ActiveFileName=${karaf.data}/log/error.log
log4j.appender.out.layout=org.apache.log4j.PatternLayout
log4j.appender.out.layout.ConversionPattern=%d{yyyy-MM-dd} %d{HH:mm:ss}  | %-5.5p | %-16.16t | %-32.32c{1} | %X{bundle.id} - %X{bundle.name} - %X{bundle.version} | %m%n
log4j.appender.out.append=true

# CXF request and response info:
# * ERROR - none
# * INFO - just headers (default)
# * DEBUG - whole, with payloads
log4j.additivity.com.cumulocity.rest.interceptors=false
log4j.logger.com.cumulocity.rest.interceptors=INFO,access

# Access appender
log4j.appender.access=org.apache.log4j.rolling.RollingFileAppender
log4j.appender.access.rollingPolicy=org.apache.log4j.rolling.FixedWindowRollingPolicy
log4j.appender.access.rollingPolicy.maxIndex=10
log4j.appender.access.triggeringPolicy=org.apache.log4j.rolling.SizeBasedTriggeringPolicy
log4j.appender.access.triggeringPolicy.MaxFileSize=262144000
log4j.appender.access.rollingPolicy.FileNamePattern=${karaf.data}/log/access-%i.log.gz
log4j.appender.access.rollingPolicy.ActiveFileName=${karaf.data}/log/access.log
log4j.appender.access.layout=org.apache.log4j.PatternLayout
log4j.appender.access.layout.ConversionPattern=%d{yyyy-MM-dd} %d{HH:mm:ss}  | %-5.5p | %-16.16t | %-32.32c{1} | %X{bundle.id} - %X{bundle.name} - %X{bundle.version} | %m%n
log4j.appender.access.append=true

# Error response info:
# * INFO - just error message (default)
# * DEGUB - full stack trace
log4j.logger.com.cumulocity.rest.mediatypes=INFO

Change the following entries to adjust the log levels:

log4j.rootLogger=INFO,out,osgi:*
		
log4j.logger.com.cumulocity.rest.interceptors=INFO,access
		
og4j.logger.com.cumulocity.rest.mediatypes=INFO

Adjust the log levels by changing the level attribute according to the following values. The levels are inclusive - meaning a given level will also include all “lower” log levels, e.g. when you set the level to WARN you will also get ERROR events.

Save the file. It is re-read by the application every few minutes so you do not have to restart the Java process.

Diagnostics

The diagnostic utility is enabled by default and runs periodically. However, this can also be triggered manually on demand. To execute it manually, follow the steps below.

cd opt/c8y/utilities/diagnostic-utility
sudo ./run_data_collector.py

Hardware information

The basic hardware information of the target system is captured. These reports are placed under the ‘hardware’ directory.

The following hardware information is available:

Software information

The basic software information of the target system is captured. These reports are placed under the ‘software’ directory.

The following software information is available:

Cumulocity information

This section contains information on the running Cumulocity processes, health endpoint check result, Cumulocity logs etc.

The following Cumulocity information is collected:

Jstat dumps can provide information on the following options:

• class - Statistics on the behavior of the class loader
• compiler - Statistics of the behavior of the HotSpot Just-in-Time compiler
• gc - Statistics of the behavior of the garbage collected heap
• gccapacity - Statistics of the capacities of the generations and their corresponding spaces
• gccause - Summary of garbage collection statistics (same as -gcutil), with the cause of the last and current (if applicable) garbage collection events
• gcnew – statistics of the behavior of the new generations
• gcnewcapacity - Statistics of the sizes of the new generations and its corresponding spaces
• gcold - Statistics of the behavior of the old and permanent generations
• gcoldcapacity - Statistics of the sizes of the old generations
• gcpermcapacity - Statistics of the sizes of the permanent generations
• gcutil – Summary of garbage collection statistics
• printcompilations – HotSpot compilation method statistics

As per the current implementation, the 5 statistics counts are collected at an interval of 50ms. To improve performance, this task is executed via threads.

Optional startup parameters

Following are the supported startup parameters for the monitor and the data collector.

Data collector

The data collector can be started by running the “run_data_collector.py” script located under “/opt/c8y/utilities/diagnostic-utility/src/”.

Following are the supported command line arguments which can be used while invoking the script. More than one of the supported arguments can be used simultaneously.

• -h or –hardware: Allows the script to collect only the hardware information
• -s or –software: Allows the script to collect only the software information
• -c or –cumulocity: Allows the script to collect only the cumulocity information

Monitor

The monitor can be started by running the “run_monitor.py” script located under “/opt/c8y/utilities/diagnostic-utility/src/”.

The monitor script supports only one optional startup parameter:

• -s or –skipDataCollector: Allows the user to skip the data collection even if one or more monitored components is not working.

Utility configuration file

The diagnostic utility can be customized using a properties file located under “/etc/diagnostic-utility/diagnostic_utility.properties”.

Following are the available keys used in the configuration file:

Upgrading the system

Info: If you use opcua-agent-server and you are upgrading from an earlier version of Edge to the latest version, you must execute the below mentioned commands in your EDGE VM to backup opcua-agent-server data. This has to be performed before following any of the steps described below.
Additionally, please request the Cumulocity support for root user credentials if you don’t have it.

mkdir -p /usr/edge/properties/opcua/db
cp -a /etc/opcua/db/. /usr/edge/properties/opcua/db/

Upgrading the system on VMware ESXi

To upgrade the Edge VM on VMware ESXi:

1. Power off the Edge VM from ESXi console.

2. Select the Edge VM and click Actions > Export.

3. Click OK to download the files.
   This action downloads the following files:

   • disk-0.vmdk (primary disk)
   • disk-1.vmdk (secondary disk)
   • An OVF file (for example, Generic-Edge-10.4.0.0.8.ovf)

4. Download the latest version to a separate folder.
   For example, if you are upgrading to Cumulocity IoT Edge version 10.5, save the following files to a separate folder:

   • EDGE-server-10.5.0.0-147-ESXi-disk1.vmdk
   • EDGE-server-10.5.0.0-147-ESXi-disk2.vmdk
   • EDGE-server-10.5.0.0-147-ESXi.ovf
   • EDGE-server-10.5.0.0-147-ESXi.mf

5. Copy the DATA/secondary disk of the exported VM to the latest version.
   For example, copy the secondary disk disk-1.vmdk to the folder to which you have downloaded the latest version.

6. Remove the secondary disk of the latest version and rename the secondary disk of the exported version to the latest version.
   For example:

   • Delete EDGE-server-10.5.0.0-147-ESXi-disk2.vmdk
   • Rename disk-1.vmdk to EDGE-server-10.5.0.0-147-ESXi-disk2.vmdk.

7. If the size of the secondary disk has expanded, you must update the OVF file of the latest version accordingly. In the OVF file, search for the fields ovf:capacity and ovf:capacityAllocationUnits and update the values as per the values in the exported OVF file.

8. Import the Edge VM image as described in Setting up ESXi.

Upgrading the system on VirtualBox

To upgrade the Edge VM on VirtualBox:

1. Power off the VM. To do so, access the VM and run the following command in the terminal:
   

sudo shutdown

2. Navigate to the VM directory. For VirtualBox it is [USER-DIRECTORY]\VirtualBox VMs. Move the entire VM folder (e.g. EDGE-server) to a safe location.

3. In the VirtualBox Manager, right-click on the VM and then click Remove. In the upcoming window, select Remove Only so that the VM will not be deleted.

4. Import the new version of EDGE server. To do so, click File and then Import Appliance in the VirtualBox Manager. Select the new OVA file (upgraded one supplied by Cumulocity) and click Open to import it.
   The new Edge version should now be visible in the VirtualBox Manager in powered-off state.

5. Right-click on the VM and select Settings. In the upcoming window, select Storage. Two disks should be visible called “disk001.vmdk” and “disk002.vmdk”.

6. Right-click on the second disk i.e. disk002.vmdk, and select Remove attachment.

7. Right-click on the second controller i.e. SATA controller and select Add Hard Disk. In the upcoming window, select Choose existing disk.

8. In the explorer, navigate to the previously backed-up EDGE folder as per step 2. Select the secondary disk, i.e. disk002.vmdk, and click OK.
   

9. Power on the VM by clicking the Start button.

10. Invoke the post-installer by running the script post-installation.sh located in the directory /opt/c8y/utilities.
    

sudo ./post_installation.sh

11. Choose the post-upgrade task (option 4). Provide inputs if prompted and complete the post-upgrade task. For details see Configuration.

12. Verify, whether DB and property files are intact and unchanged.

Backup and restore

Backup

To create a backup, follow the steps below.

1. To power off the VM, right-click on the respective VM (e.g. EDGE-server) in the VirtualBox Manager and click Close > Power off.
   
   
   
2. Click File at the top right and then Export Appliance.
   
   
   
   
3. Enter the location and name as desired and click Export.
   
   

Info: You may append the current date to the image name (e.g. EDGE-server-12-06-2018.ova) so as to have a track of the backups.

Restore

To create a restore, follow the steps below.

1. In the VirtualBox Manager, click File at the top right and then Import Appliance.
   
   
   
2. In the upcoming window, browse for the OVA image and select it.
   
   
   
3. Power on the VM by clicking on the Start button.

Administration and configuration

Unlocking the tenant admin user

The tenant admin user could be locked if incorrect credentials are passed during login into UI, REST API or MQTT.

To unlock the tenant admin user, perform the following steps:

1. Login as sysadmin user.
2. In the Administration application, navigate to the Users page and open the tenant admin user.
3. Activate the user account by switching the toggle next to the username to Enabled and save your settings.

Configuring email server and password template settings

To configure the “reset password” template and email server settings, perform the following steps:

1. Log into the management tenant using https://<Edge_VM_IP_Address>/apps/administration/index.html#/configuration.
   • Tenant: management
   • User: edgeadmin
   • Password: Will be the same as the Edge tenant admin password provided in the post-installation execution
     
2. Update the email server details and templates following the instructions in Administration > Changing settings> Configuration settings in the User guide.

Increasing the system performance

If the system performance is slow, the memory should be increased. First, increase the memory of the VM. This is done by stopping the VM and increasing its memory.

Increasing the VM memory should be followed by a JVM memory increase.

This is done by starting the VM (follow the steps described in the Installation section). Log into VM, open the file /usr/share/cumulocity-core-karaf/bin/setenv and edit it as described below. The parameter is the following, default size is 1024.

export JAVA_MAX_MEM=1024M # Maximum memory for the JVM

After increasing the size, restart Karaf by executing:

[admin@server ~]$  sudo service cumulocity-core-karaf stop

and

[admin@server ~]$  service cumulocity-core-karaf start

Log rotation

The components for which log rotation is configured are as follows:

The components for which log rotation is configured are as follows:

For microservices, there currently is no specific log rotation configured.

Time synchronization

For many use cases, and especially when using APAMA, time synchronization must be available, i.e. the time inside the VM must be synchronized with the time of the host OS and with devices sending data.

Out of the box, for VMWare-based installations, vmtools is responsible for time synchronization with the host OS. For VirtualBox-based installations, VirtualBox guest additions is responsible.

Additionally, chrony or ntp services can be configured by end users based on their time synchronization needs. Refer to the respective documentation for the configuration of these services. These services are by default stopped and disabled in Edge and can be enabled by standard commands.

Restarting services

Karaf processes

On the Edge server, run the following commands as admin user.

opcua-agent-server

On the Edge server, run the following commands as admin user.

smartrule

On the Edge server, run the following commands as admin user.

Apama

On the Edge server, run the following commands as admin user.