How to troubleshoot Ethernet Network Ports in ONTAP using ifstat
Applies to
Description
- This article contains a list of most network port operational and troubleshooting workflows. However, it is not a comprehensive list.
- This can be used to narrow your search to the more commonly utilized troubleshooting KB articles, broken down to a specific category.
- The primary tool to identify network port issues in ONTAP is the “ ifstat” command, available on both Data ONTAP 7-mode filers, as well as from the node-shell of a Data ONTAP 8.3.x or ONTAP 9 cluster.
Note: Understanding the different shells for CLI commands (cluster administrators only)
For a more detailed view of the 'ifstat' command, please see the manual page, located here.
NAME
na_ifstat - Displays device-level statistics for network interfaces.
SYNOPSIS
ifstat [ -z ] -a | interface_name
DESCRIPTION
- The ifstat command displays statistics about packets received and sent on a specified network interface or on all network interfaces.
- The statistics are cumulative from the time when the node was booted, or the counter was reset
- The -z argument clears the statistics.
ifstat -z e0a
clears the statistics for port e0a onlyifstat -z -a
clears the statisticts for all ports
ifstat -a
displays statistics for all network interfaces including the virtual host and the loopback address.ifstat e0a
displays statistics for port e0a only
HA CONSIDERATIONS
- In takeover mode, the ifstat command displays combined statistics about packets processed by the local network interface and packets processed by the local network interface on behalf of the network interface on the failed node.
- The statistics displayed by the ifstat command are cumulative. That is, a giveback does not cause the ifstat command to zero out the statistics.The counters returned by ifstat for any given Network Interface Port can vary by version of ONTAP or NIC Hardware Provider.
Example ifstat output in ONTAP 9.2 on a Quad Gigabit Ethernet Controller 82580:
::*> system node run -node cm2520-ks6-01 -command ifstat e0a
-- interface e0a (3 hours, 45 minutes, 32 seconds) --
RECEIVE
Total frames: 0 | Total bytes: 0 | Total errors: 0
Total discards: 0 | Multi/broadcast: 0 | Non-primary u/c: 0
CRC errors: 0 | Runt frames: 0 | Fragment: 0
Long frames: 0 | Jabber: 0 | No buffer: 0
Xon: 0 | Xoff: 0 | Jumbo: 0
Noproto: 0 | Error symbol: 0 | Illegal symbol: 0
Bus overruns: 0 | LRO segments: 0 | LRO bytes: 0
LRO6 segments: 0 | LRO6 bytes: 0 | Bad UDP cksum: 0
Bad UDP6 cksum: 0 | Bad TCP cksum: 0 | Bad TCP6 cksum: 0
Mcast v6 solicit: 0
TRANSMIT
Total frames: 0 | Total bytes: 0 | Total errors: 0
Multi/broadcast: 0 | Max collisions: 0 | Single collision: 0
Multi collisions: 0 | Late collisions: 0 | Xon: 0
Xoff: 0 | Jumbo: 0 | Cfg Up to Downs: 0
TSO non-TCP drop: 0 | Split hdr drop: 0 | Timeout: 0
TSO segments: 0 | TSO bytes: 0 | TSO6 segments: 0
TSO6 bytes: 0 | HW UDP cksums: 0 | HW UDP6 cksums: 0
HW TCP cksums: 0 | HW TCP6 cksums: 0 | Mcast v6 solicit: 0
DEVICE
Mcast addresses: 1 | Rx MBuf Sz: 4096
LINK INFO
Speed: 0 | Duplex: full | Flowcontrol: full
Current state: down | Up to downs: 0
- Some of the more common counters used to diagnose network issues from ifstat output are highlighted in red above.
IMPORTANT
What do these counters mean?
- Recognize these counters having any particular value is only particularly meaningful if you know how long the counters have been running and the rate they are incrementing.
- What may appear to be a large number of errors may not indicate a significant problem if they counters have not been reset (by rebooting or executing ifstat -z).
- Alternatively, we can periodically check these counters to see how rapidly they are incrementing.