Cloud Latency Monitor Plug-in

You must be a MyPeek member to download our extensibility solutions. Click here to learn more about the benefits of membership and find out how to sign up for free.

Download Plug-In

You must have a valid maintenance contract to download this file.
If you are not a Savvius Maintenance Customer but would like to purchase a Maintenance contract for your Savvius product please click here for sales information.
Downloaded : 854 Times
Rating : 5 of 5

View Comments (4)

Category : Plug-Ins

Submitted By : Victor Wu

Important: This extensibility solution supports both 32-bit and 64-bit installations of Omnipeek Professional or Enterprise

WildPackets Cloud Latency Monitor for OmniPeek (CLM) is WildPackets’s latency monitor tool with the intent to assist in monitoring cloud activities.

From the following figure, we can compare the latency patterns between google ICMP and HTTP requests. We can also compare google HTTP and Yahoo HTTP requests.


In this V 2.0 release, the CLM comes with the following features:

1. Real-time monitoring capability for ICMP (ping) and HTTP (GET)

2. User can add/delete filters

3. Ping scan on IP range

4. System Log option

1. User can specify a SysLog server as logging/alarm purpose

5. Latency statistics

1. The following information can be specified by user

1.Filter Name, such as google

2.Target Hostname or IP Address, such as, or

1. If a host name is entered, it will be resolved into an IP address

3.The resolved IP address varies from time to time due to the nature of DNS

4.For ICMP request:

1. Data to be sent out with ping

§ Can be specified as size of data in bytes between 18 – 1000

§ An ICMP socket requires at least 18-byte user data

§ Data field of the ICMP packet will be filled with 0x10

2. Or as data string by itself

§ Data size should be specified in INTEGER, "99.9" will be treated as raw data

§ If data size is less than 18 bytes, the data will be padded with 0x00 at the end

3. Quotation marks can be used to quote a string like "0123", the quotation marks at the two ends will be removed during packing the frame

§ This is useful when raw data "10" is desired rather than the buffer size 10.

5. For HTTP request:

1. An HTTP 1.0 GET request will be sent out to the target IP address.

2. Latency is measure as the time between making request and the time the transmission of the whole page is completed.

6. Alarm threshold

1. If the responding latency is larger than the specified threshold, system log/alarm will be provided if SysLog server is specified

2. The following is window-based meaning their calculation is based on window (i.e., collection) size of the samples, default to 1000 samples

1.Highest Latency

2.Lowest Latency

3.Average Latency

4.Loss percentage

3. Sent and Received pings, these are the overall counts

6. Real-time graph on latency (in ms)

1. Should have all active filters in different colors

7. Report (HTML format), including

1. Statistics of all active filters

2. Graph


How to use buttons:
This is the toolbar of CLM.

Button 1: START a new monitoring session ( only in standalone version, not available in CLM PlugIn )

Button 2: STOP a monitoring session ( only in standalone version, not available in CLM PlugIn )

Button 3: SysLog option: user can specify a server which will be noticed when there is a latency greater than what user set.

Button 4: Ping Scan: User can specify a subnet range, CLM will scan through the sub network for 20 seconds and add those reachable IPs as new filters.

Button 5: Add Filter: user can create new monitoring target by specifying a name, hostname or IP, send buffer or data, and Alarm threshold. User need to specify the request as ICMP or HTTP.

Button 6: edit an existing filter

Button 7: duplicate an existing filter (and give a new name)

Button 8: delete a highlighted filter

Button 9: import pre-defined filters from xml file

Button 10: export filters to xml file

Button 11: select all filters

Button 12: unselect all filters

Button 13: Save Report in HTML format

Button 14: Set the refresh or sampling frequency (1 sec/2secs/3 secs ...)

Button 15: zoom in the graph

Button 16: zoom out the graph

PopUp Menu

Delete All” pop-up menu
  • Let you delete all the filters at once

v2.0 Limitations

  1. The time range of the graph will be limited to 86400 points, which is 86400 seconds (24 hours)
    • If we reach the point that more 86400 samples are processed, the graph will be reset and start with a brand new session
  2. Due to the nature of the DNS, a host name may not always map to the same IP address for each new monitoring section
    • may be resolved as or or something else
  3. Currently support HTTP/1.0 GET for HTTP service monitor


Version 2.0 04/17/2012
- added HTTP service monitor capability

Version 1.0 03/22/2012
- renamed to Cloud Latency Monitor
- Better graph
- Report capability

Version 2/22/2009
- Fixed performance bug
- Added alarms
- Added Delete all

Version 2/10/2009
- Added Scan button to toolbar
- Fixed lots of bugs
- Legend works better

Version 2/9/2009
- Built-in Graph
- Hi-Def Toolbar
- Interval setting from 1 second to 10 minutes
- Rolling High/Low/Average values in list
- Text or size of data per entry

Version 11/21/08
- Works much better now - takes hostname or IP

Version 11/03/08
- Release to MyPeek


Posted by: Ofer Mordechay on Tuesday, February 17th, 2015 at 10:34 AM
Hi Joe! That is a great idea. You can already specify the payload of the ping packet, but it would be painful to create a 1024 byte packet by typing in 1024 characters. In the next release we are going to add an MTU feature, so you can just specify 1024 as the MTU size.

We would love to hear some suggestions from other folks out there for useful features. Otherwise we will just make some up.
Posted by: Chris Bloom on Friday, December 12th, 2008 at 9:54 AM
Is there a way to just set a MTU size?
Posted by: Joe Habib on Tuesday, December 9th, 2008 at 8:21 AM
I have found that some clients will only respond to pings with the data "1234". For example, "hello" will not get a reponse. I do not know why. Does anybody?
Posted by: Chris Bloom on Tuesday, November 18th, 2008 at 3:46 PM