Troubleshooting the RPS Components
Understand Alarm Types and Severity Levels on EX Series Switches
This topic applies only to the J-Web Application package.
Alarms alert you to conditions that might prevent normal operation of the switch. Before monitoring alarms on a Juniper Networks EX Series Ethernet switch, become familiar with the terms defined in Table 1.
Table 1: Alarm Terms
Signal alerting you to conditions that might
prevent normal operation. On a switch, the alarm signal is the
Failure event that triggers an alarm.
Seriousness of the alarm. If the Alarm (
Preset alarm triggered by a physical condition on the switch such as a power supply failure, excessive component temperature, or media failure.
Preset alarm triggered by a missing rescue configuration or failure to install a license for a licensed software feature.
Note: On EX6200 switches, a system alarm can be triggered by an internal link error.
The switch supports these alarms:
Chassis alarms indicate a failure on the switch or one of its components. Chassis alarms are preset and cannot be modified.
System alarms indicate a missing rescue configuration. System alarms are preset and cannot be modified, although you can configure them to appear automatically in the J-Web interface display or the CLI display.
Alarm Severity Levels
Alarms on switches have two severity levels:
Major (red)—Indicates a critical situation on the switch that has resulted from one of the following conditions. A red alarm condition requires immediate action.
One or more hardware components have failed.
One or more hardware components have exceeded temperature thresholds.
An alarm condition configured on an interface has triggered a critical warning.
Minor (yellow or amber)—Indicates a noncritical condition on the switch that, if left unchecked, might cause an interruption in service or degradation in performance. A yellow or amber alarm condition requires monitoring or maintenance.
A missing rescue configuration generates a yellow or amber system alarm.
Check Active Alarms with the J-Web Interface
This topic applies only to the J-Web Application package.
Use the monitoring functionality to view alarm information for the EX Series switches including alarm type, alarm severity, and a brief description for each active alarm on the switching platform.
To view the active alarms:
- Select Monitor > Events and Alarms > View Alarms in the J-Web interface.
- Select an alarm filter based on alarm type, severity, description, and date range.
- Click Go.
All the alarms matching the filter are displayed.
When the switch is reset, the active alarms are displayed.
Table 2 lists the alarm output fields.
Table 2: Summary of Key Alarm Output Fields
Category of the alarm:
Alarm severity—either major (red) or minor (yellow or amber).
Brief synopsis of the alarm.
Date and time when the failure was detected.
Monitor System Log Messages
This topic applies only to the J-Web Application package.
Use the monitoring functionality to filter and view system log messages for EX Series switches.
To view events in the J-Web interface, select Monitor > Events and Alarms > View Events.
Apply a filter or a combination of filters to view messages. You can use filters to display relevant events. Table 3 describes the different filters, their functions, and the associated actions.
To view events in the CLI, enter the following command:
Table 3: Filtering System Log Messages
System Log File
Specifies the name of a system log file for which you want to display the recorded events.
Lists the names of all the system log files that you configure.
By default, a log file, messages, is included in the /var/log/ directory.
To specify events recorded in a particular file, select the system log filename from the list— for example, messages.
Select Include archived files to include archived files in the search.
Specifies the name of the process generating the events you want to display.
To view all the processes running on your system, enter the CLI command show system processes.
For more information about processes, see the Junos OS Installation and Upgrade Guide.
To specify events generated by a process, type the name of the process.
For example, type mgd to list all messages generated by the management process.
Specifies the time period in which the events you want displayed are generated.
Displays a calendar that allows you to select the year, month, day, and time. It also allows you to select the local time.
By default, the messages generated during the last one hour are displayed. End Time shows the current time and Start Time shows the time one hour before End Time.
To specify the time period:
Specifies the event ID for which you want to display the messages.
Allows you to type part of the ID and completes the remainder automatically.
An event ID, also known as a system log message code, uniquely identifies a system log message. It begins with a prefix that indicates the generating software process or library.
To specify events with a specific ID, type the partial or complete ID— for example, TFTPD_AF_ERR.
Specifies text from the description of events that you want to display.
Allows you to use regular expressions to match text from the event description.
Note: Regular expression matching is case-sensitive.
To specify events with a specific description, type a text string from the description with regular expression.
For example, type ^Initial* to display all messages with lines beginning with the term Initial.
Applies the specified filter and displays the matching messages.
To apply the filter and display messages, click Search.
Resets all the fields in the Events Filter box.
To reset the field values that are listed in the Events Filter box, click Reset.
Generate Raw Report
Generates a list of event log messages in nontabular format.
To generate a raw report:
Note: Starting in Junos OS Release 14.1X53, a Formatted Report can be generated from event log messages being loaded in an Events Detail table. The Generate Report button appears only after event log messages are completely loaded in the Events Detail table. The Generate Raw Report button is displayed while event log messages are being loaded.
Generates a list of event log messages in tabular format, which shows system details, events filter criteria, and event details.
To generate a formatted report:
Table 4 describes the Event Summary fields.
By default, the View Events page in the J-Web interface displays the most recent 25 events, with severity levels highlighted in different colors. After you specify the filters, Event Summary displays the events matching the specified filters. Click the First, Next, Prev, and Last links to navigate through messages.
Table 4: Viewing System Log Messages
Displays the name and ID of the process that generated the system log message.
The information displayed in this field is different for messages generated on the local Routing Engine than for messages generated on another Routing Engine (on a system with two Routing Engines installed and operational). Messages from the other Routing Engine also include the identifiers re0 and re1 that identify the Routing Engine.
Severity level of a message is indicated by different colors.
A severity level indicates how seriously the triggering event affects switch functions. When you configure a location for logging a facility, you also specify a severity level for the facility. Only messages from the facility that are rated at that level or higher are logged to the specified file.
Displays a code that uniquely identifies the message.
The prefix on each code identifies the message source, and the rest of the code indicates the specific event or error.
The event ID begins with a prefix that indicates the generating software process.
Some processes on a switch do not use codes. This field might be blank in a message generated from such a process.
An event can belong to one of the following type categories:
Displays a more detailed explanation of the message.
Displays the time at which the message was logged.
Troubleshooting the EX Series Redundant Power System Power On and Power Backup Issues
This topic provides troubleshooting information for problems related to the EX Series Redundant Power System (RPS).
The EX Series RPS Is Not Powering On
Description: The RPS does not power on even though it has a power supply installed and is connected to an AC power source outlet.
Environment: The RPS with one EX-PWR3-930-AC power supply installed in it is connected to a switch.
Symptoms: The SYS LED on the power supply side of the RPS is off, and when you check the RPS status using the CLI command show chassis redundant-power-system, the message No RPS connected is displayed.
A power supply must be installed in the middle slot on the RPS to power on the RPS.
Install a power supply in the middle slot on the power supply side of the RPS and verify that the AC power source outlet is properly connected to it. See Installing a Power Supply in the EX Series Redundant Power System.
Verify that the AC OK LED and the DC OK LED on the power supply in the RPS are lit green.
A Switch Is Not Recognized by the RPS
Description: I cannot set up the RPS.
A switch must be active to be recognized by the RPS.
Activate the switch by configuring it and issuing a commit statement.
An Error Message Indicates That an RPS Power Supply is Not Supported
Description: An RPS error message indicates that an RPS power supply is not supported.
RPS supports only one power supply, the EX-PWR3-930-AC. If you install another similar power supply, it may fit in the slot but it is not compatible with RPS.
The power supply shipped with your RPS (in a separate box) is an EX-PWR3-930-AC. If you installed more power supplies, you ordered them separately. Replace any other power supply model (such as the EX-PWR2-930-AC) with an EX-PWR3-930-AC model.
The EX Series Redundant Power System Is Not Providing Power Backup to a Connected Switch
Description: The RPS does not provide power backup to a connected switch.
Environment: The RPS has an EX-PWR3-930-AC power supply installed in the middle power supply slot and is connected to two switches with power loss, one connected to RPS switch connector port 1 and the other on port 2.
Symptoms: The status LED on the associated switch connector port is not blinking green–it is either solid green (connected) or not lit (off).
The RPS provides backup power based on the power priority assigned to each switch.
If the status LED on a switch connector port is off, ensure that the RPS cable is properly connected to both the RPS and the switch, and ensure that the priority configured for the switch is not 0. See show redundant-power-system status.
If the status LED on switch connector port 1 is on and is steadily green, check the backup priority configured for the switch and assign it a higher priority. See Determining and Setting Priority for Switches Connected to an EX Series RPS
If the status LED on switch connector port 1 is amber, check if the RPS has enough power supplies installed in it to provide backup power. if it does not, install a power supply in an empty power supply slot on the RPS. See Installing a Power Supply in the EX Series Redundant Power System.
If the status LED on switch connector port 1 is still off, check the priority configured for the switch. Ensure that the is not set to 0, which means off. See show redundant-power-system status. The priority assigned must be from 1 through 6. See Determining and Setting Priority for Switches Connected to an EX Series RPS.
Verify that a dedicated power supply is installed in the switch. The RPS cannot boot a switch that does not have a dedicated power supply. See Installing a Power Supply in the EX Series Redundant Power System.
Also keep in mind that when the command request redundant-power-system multi-backup has been set, support for switches that supply PoE is not guaranteed. To reverse this setting, use the command request redundant-power-system no-multi-backup.
The Wrong Switches Are Being Backed Up
Description: Four or more switches are connected to an RPS with three power supplies. When all four switches fail, the wrong three switches have .
Environment: Four or more switches are connected to an RPS with three power supplies. One or more switches provide PoE to other devices.
Symptoms: When all four switches fail, the wrong three switches have .
The RPS provides backup power based on the power priority assigned to each switch. This is derived from two configurations, one of which has precedence over the other one. Initial is derived from the location of the port used to attach a switch—the leftmost connector has lowest priority and the rightmost connector has highest priority. The second, dominant priority configuration is derived from a CLI priority setting on the switch itself. With this CLI configuration, 6 is highest priority and 1 is the lowest priority.
Connect the three switches to the three rightmost connectors on the RPS. Then, using the CLI on each switch, set each switch’s priority to 1 using the redundant-power-system configuration command redundant-power-system 1. Now, physical connection location is determining .
If you do not want to change the cabling on the switches, you can use the configuration statement redundant-power-system on all four switches, assigning priority 6 (highest), 5, 4 and 3 to the appropriate switches. Priority configuration on the switch always overcomes set by connector location.
Six Switches That Do Not Require PoE Are Not All Being Backed Up
Description: Only three switches out of six are simultaneously backed up when all switches experience power supply failure. None of these switches supply PoE power to any device.
Environment: The RPS with three EX-PWR3-930-AC power supplies installed in it is connected to six switches, none of which is connected to a non-PoE device.
Symptoms: Only three switches out of six are simultaneously backed up when all switches experience power supply failure. None of these switches supply PoE power to any device.
Each power supply can support two switches that do not need enough power for PoE, as long as you configure the RPS to do so.
From any of the attached switches, issue the request redundant-power-system multi-backup command from the operational mode. Now standard power will be supplied to two non-PoE switches per power supply.