cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1225
Views
0
Helpful
5
Replies

Questions about output drops on an interface

burleyman
Level 8
Level 8

I am getting the following when I do a show interface. What does "Total output drops" mean and should I be worried?

Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 53

Queueing strategy: Class-based queueing

Output queue: 0/1000/64/53 (size/max total/threshold/drops)

Thanks,

Mike

1 Accepted Solution

Accepted Solutions

mmorris11
Level 4
Level 4

Output drops indicate that your interface had become overwhelmed with outbound traffic. Whether or not to be worried depends on the frequency of drops; in the "sh int" command you can use the packet count with this information to calculate the packet drop ratio. If it is in the low single digits, I wouldn't sweat it. If it is more than that, you might need to look into your traffic management policies and come up with a strategy that avoids dropping important traffic. This often includes getting more bandwidth ;)

HTH pls rate!

View solution in original post

5 Replies 5

mmorris11
Level 4
Level 4

Output drops indicate that your interface had become overwhelmed with outbound traffic. Whether or not to be worried depends on the frequency of drops; in the "sh int" command you can use the packet count with this information to calculate the packet drop ratio. If it is in the low single digits, I wouldn't sweat it. If it is more than that, you might need to look into your traffic management policies and come up with a strategy that avoids dropping important traffic. This often includes getting more bandwidth ;)

HTH pls rate!

Thanks that helped. My drops were 53 output drops for 860036 output packets so it looks like I should be fine.

Mike

How we solve it, how much i need to increase the hold-quee ???

Disclaimer

The Author of this posting offers the information  contained within this posting without consideration and with the  reader's understanding that there's no implied or expressed suitability  or fitness for any purpose. Information provided is for informational  purposes only and should not be construed as rendering professional  advice of any kind. Usage of this posting's information is solely at  reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever  (including, without limitation, damages for loss of use, data or  profit) arising out of the use or inability to use the posting's  information even if Author has been advised of the possibility of such  damage.

Posting

"Solving" drops can be quite complex.  Sometimes "solving" doesn't totally eliminate drops, but "solving" often does minimize the number of drops.

Drops can be used for TCP flow control.  When used in this role, the idea is to drop the minimum number of packets to have the flow's source send at about the available bandwidth.

Increasing queue depth often, but not always, will mitigate drops but it also can increase latency.  I.e. "solving" drops might warrant decreasing queue depth, although not often done on Cisco devices since their default queue depth usually isn't too large.

For TCP, optimal queue size should support BDP (the bandwidth delay product).  Too small a queue leads to premature drops, too large a queue leads to needless queuing latency.

Disclaimer

The Author of this posting offers the information contained within this posting without consideration and with the reader's understanding that there's no implied or expressed suitability or fitness for any purpose. Information provided is for informational purposes only and should not be construed as rendering professional advice of any kind. Usage of this posting's information is solely at reader's own risk.

Liability Disclaimer

In no event shall Author be liable for any damages whatsoever (including, without limitation, damages for loss of use, data or profit) arising out of the use or inability to use the posting's information even if Author has been advised of the possibility of such damage.

Posting

BTW, I recall Cisco recommending only being concerned when drop percentage exceeds 1%, although I don't recall whether this was per hop or end-to-end (the latter makes the most sense).  This recommendation didn't, I believe, take into account different kinds of traffic, for example even 1% might be too high for VoIP but a bulk data transfer using TCP could probably go a little higher.