Measurement strategy
The measurement strategy for most permanent deployments is to capture samples eight times a day for a defined length of time (currently 90 seconds). The sampling interval was originally set to 2 minutes, but has since been lowered to 90 seconds in order to limit the amount of data being captured each day.
We also have sites now which have longer sampling intervals, captured just twice a day. We currently collect between 1.5 and 3.2 Gigabytes of compressed data each day.
The The last column statistics reflect the total data collected from that site on the date listed above the table, with the frequency of sampling in parentheses. The frequency of sampling is also the number of data files available for each day.
We also have special traces and new deployments in locations which complement existing monitors and allow for multi-hop instrumentations (several monitors along a communications path). Please see the Special Traces overview page.
Legend
| CHDLC |
Cisco HDLC, a lightweight header of one byte, one spare and an EtherType field |
| CHDLC multi |
a particular configuration of CHDLC allowing multiplexed channels to coexist on a link |
| DAG3.2 |
monitor is capturing with
DAG3.2
OC3c/OC12c ATM/PoS cards |
| FORE |
monitor is capturing with Coral-1.1 and FORE SBA-200E ATM cards |
| LLC/SNAP |
encapsulation as per RFC1483,
Section 4.1, Payload Format for Routed IP PDUs |
| POINT |
monitor is capturing with Coral and AppTel POINT OC12c PoS cards |
| PoS |
Packet-over-SONET |
| PPP |
Point-to-Point Protocol as in RFC1662 |
| none (VC multi) |
encapsulation as per RFC1483,
Section 5.1. VC Based Multiplexing of Routed Protocols |
| *** |
site not responding on date shown at top of column |
| 
