SNMP

For Release 2.0.2, Media Flow Controller supports the Entity and Asset SNMP MIBs for discovery, asset management, alarms, and traps; that is HOST-REOURCES, HOST-RESOURCES-TYPES, IF, IP, IDP, and TCP MIBs.

Typically, SNMP uses UDP (User Datagram Protocol) ports 161 for the agent and 162 for the manager. The Manager may send requests from any available port (source port) to port 161 in the agent (destination port). The agent response is given back to the source port. The Manager receives traps on port 162. The agent may generate traps from any available port. See Table 17, below for information on SNMP traps, possible causes, and recommended actions.

Note! At this time, Media Flow Controller does not support provisioning over SNMP V3.

 

Table 17 SNMP Alarms, Possible Causes, and Recommended Actions  

Event Name (MIB/ stat Name)	Cause	Action
System Traps
Interface UP/Down (cpu)	The network interface on Media Flow Controller went operationally UP or DOWN. This event could happen because a wire was unplugged, faulty cable, loose cabling, or network configuration issue. This event causes high-availability to kick in and requests may be forwarded to another Media Flow Controller or POP.	Check the network cables that go into the NIC card and make sure that all cabling is OK. Check the network configuration between the switch/router and server. Make sure that the interface speeds match on the switch and router because this mismatch could cause speed auto-negotiation problems.
Process Crashed (procmgr)	The Media Flow Controller process can restart if there is a software problem. A core file is generated. This event causes high-availability to kick in and requests may be forwarded to another Media Flow Controller or POP. The crashed Media Flow Controller process should restart very quickly after which new connections/requests are served. If the core Media Flow Controller server process does not crash, requests are served in a normal manner. For example, if the management process restarts, requests are not affected.	Escalate to Juniper Networks Support to look at the core file. However, request processing may proceed automatically because the Media Flow Controller process restarts on its own.
Process Got Stuck (procLivenessFailure)	The Media Flow Controller is not serving out any more requests. This event causes high-availability to kick in and requests are forwarded to another Media Flow Controller or POP. This Media Flow Controller may not be able to recover.	Processing may proceed automatically because the Media Flow Controller process restarts on its own; if it doesn’t, escalate to Juniper Networks Support to diagnose the issue.
Process Not Running (procExit)	The Media Flow Controller process is not serving out any more requests. This event causes high-availability to kick in and requests are forwarded to another Media Flow Controller or POP. This Media Flow Controller may not be able to recover.	Processing may proceed automatically because the Media Flow Controller process restarts on its own; if it doesn’t, escalate to Juniper Networks Support to diagnose the issue.
Unexpected Shutdown (unexpectedShutdown)	The server may unexpectedly shutdown due to power reasons. This is a highly unlikely event. This event causes high-availability to kick in and requests are forwarded to another Media Flow Controller or POP. This Media Flow Controller may not be able to recover.	Check all power cables. Restart Media Flow Controller and all process and configurations restart normal operation.
Hardware Resources Traps
Disk Failure Warning (smartError)	One or more disks are issuing warnings that a disk sector was corrupted, un-writeable, or unreadable. Media Flow Controller still serves out requests from other disks. Note: If the ROOT disk is corrupted, the system may not be functional and will need to be looked at.	Escalate to Juniper Networks Support to diagnose the issue.
CPU Utilization High (cpuUtilHigh)	The CPU utilization may be high due to increased processing of requests, disk I/O, or some error. When CPU utilization is high, Media Flow Controller performance may be compromised. This event may be transitory or permanent.	If high CPU utilization is permanent, escalate to Juniper Networks Support to look at the problem.
Disk Space Low (diskSpaceLow)	Disk space low may be reported on the ROOT disk because logs may be taking too much of disk space. This may cause the ROOT disk to un-writeable.	Upload access log, cache log, error logs to a networked server and delete unneeded logs from the Juniper Networks server.
Paging High (pagingActivityHigh)	Memory utilization may be high on the Media Flow Controller server. When this happens, Media Flow Controller performance may be compromised.	Escalate to Juniper Networks Support to look at the problem.
Disk I/O High (disk-io-high)	Disk I/O per second has risen above 5 MB. Indicates that more content is served from disks.	Check if the amount of hot/popular content is more than the available RAM size. If the RAM sizing is inadequate, consider increase the RAM in the system.
Memory Usage High (memusage-high)	Memory usage has risen above 90%	Check if a single process is hogging the system memory. If so, restart the process.
Network Usage High (netusage-high)	Network utilization has risen too high	Check if Media Flow Controller is constantly fetching content from origin server. If so, verify the cache-ability parameters on the origin server (such as Cache-Control, Expiry, Last Modified Date).   Check if Media Flow Controller is trying to upload access logs to an unavailable server.
Cache Resources Traps
Bandwidth Limit (total_byte_rate)	Configured bandwidth limit has crossed on this server	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller
Cache Bandwidth (cache_byte_rate)	Current Cache bandwidth usage is too high	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller
Origin Bandwidth (origin_byte_rate)	Current Origin bandwidth usage is too high	Check if disk cache and RAM cache are properly utilized by monitoring the dashboard. Media Flow Controller may be serving lots of long-tail content or the content couldn't be accommodated in the RAM/disk cache. Check if the available disk/RAM cache sizing is adequate for the traffic load.
Disk Bandwidth (disk_byte_rate)	Current Disk BW Usage is too high	Check if the amount of hot/popular content is more than the available RAM size. If the RAM sizing is inadequate, consider increase the RAM in the system.
Averaged Cache Bandwidth (avg_cache_byte_rate)	Cache Average bandwidth usage, since the system uptime, is too high	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller.
Averaged Origin Bandwidth (avg_origin_byte_rate)	Origin Average bandwidth usage, since the system uptime, is too high	Check if disk cache and RAM cache are properly utilized by monitoring the dashboard. Media Flow Controller may be serving lots of long-tail content or the content couldn't be accommodated in the RAM/disk cache. Check if the available disk/RAM cache sizing is adequate for the traffic load.
Connection Rate (connection_rate)	Incoming connection rate is too high	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller.
Transaction Rate (http_transaction_rate)	HTTP transaction rate is too high	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller.
Port Bandwidth (perportbyte_rate)	Network port bandwidth usage is high	Configure traffic redirection thresholds in the external router/load balancer to reduce the number of requests redirected to the Media Flow Controller. Configure additional network ports in Media Flow Controller to redistribute the requests evenly across the available ports.

Configuring SNMP consists of setting a system contact and location, admin user parameters, and Trap Sinks. Note that traps are only sent if there are Trap Sinks configured and enabled. For more information on SNMPv2, see RFC 2578.

In the Management Console, use the Setup > SNMP page to do this.

In the CLI, use these commands:

snmp ? (to set all SNMP parameters) and
email notify ? (to determine who should get notifications) See “Email Notification Options” on page 65 for details.