Tier 1 carrier performance report: November, 2017

The report covers broadly used Tier 1 carriers performance results in the US and the general spread for the month of November 2017. For the US results, rare datapoints from distant locations e.g. Alaska have been omitted.

The presented analysis is based on more than 763 million successful probes that span the entire month. All data is aggregated per carrier on a daily basis and accounts for many thousands of successful probes. A control group (labeled C) is used as a base of comparison. The control group aggregates the average for all transit providers in a network, including Tier 1 carriers.

Fig. 1. Average Loss and Latency in November 2017
The charts include a control group C (gray) to allow cross comparison.

The values for October 2017 are included for cross comparison.

Fig. 2. Average Loss and Latency in October 2017
The charts include a control group C (gray) to allow cross comparison.

Average packet loss analysis:

• NTT, Level 3 and Zayo occupied the leading positions in November 2017. XO and Cogent results for November are close to the control group level;
• A higher level of average loss has been registered for GTT and Telia;
• Comparing the average packet loss of the Tier 1 carriers for the month of November with the registered results in October, ascertain for all carriers the average packet loss increased, with the exception of Hurricane Electric which kept the same level of loss as in the previous month.

Average latency analysis:

• In November, GTT has registered the highest average latency between the analyzed carriers, followed by Hurricane Electric. Level 3 has registered the lowest average latency for the analysed group, followed by NTT. XO, Telia, and Zayo results are near the control group level;
• As in October, GTT’s latency was one of the highest, followed by Hurricane Electric and Cogent. Centurylink is positioned between the poor performing Tier 1s and the control group level. The control group has registered approximately the same average latency level as in the previous month.

The charts below illustrate the performance of each carrier in comparison to the control group.

Fig. 3. Better or worse Loss and Latency in November 2017
The numbers are differences from average control group.

Fig. 4. Better or worse Loss and Latency in October 2017
The numbers are differences from average control group.

In comparison with the control group for the month of November, 2017 :

• NTT and Level 3 showed better results in terms of Loss. The leader of this group, NTT, registered a lower packet loss average than the control group level (2,78%) by ~ 0.6%, followed by Level 3 (~ 0.3%). Zayo has shown approximately the same result as the control group. GTT showed substantially worse result in terms of Loss, exceeding the packet loss average by 0,9%. Telia surpassed the packet loss average by ~ 0.6%;
• Level 3 and NTT presented better results in terms of Latency. Level 3 latency was lower than the average latency of the control group (126,2 ms) by ~ 10 ms, and NTT by ~ 8 ms. GTT’s average latency surpassed the control group’s one by ~ 19 ms. Hurricane Electric was highlighted with a latency surplus of ~ 14 ms compared to the control group average.

In comparison with the data from October, 2017 :

• NTT and Level 3 kept their leader positions as in October in terms of Loss. XO increased the packet loss average in comparison with the previous month. Hurricane Electric improved its results in comparison with the October ones. Telia and Cogen kept the same tendency as in October, still remained in the group of Tier 1’s with higher packet Loss than the control group average;
• Level 3 and NTT remained in the winning group in terms of Latency, while Centurylink migrated into the losers’ category. GTT has registered the worse Latency result as in October.

Loss

For the Loss analysis we use a scatter plot, where average values by control group are assumed on the diagonal while the horizontal and the vertical axis highlight carrier metrics. All data-points below the diagonal represent the better performing carriers and vice versa.

Abnormally large losses are still registered for a large number of datapoints. We consider excessive an average above 4.5% packet loss. Given the fact that Tier 1 carriers are characterized by both low loss values for some networks and abnormally high losses for other networks, the conclusion is that high loss values are not caused by the carriers themselves but rather are caused by the networks they service or the networks they peer with. Whether the true cause is poor design, over-provisioned links or deficiencies in peering governance – this report cannot tell. What we can mention is that for many networks, whether permanently or sporadically, there is definitely an opportunity to improve things.

Fig. 5. Loss values spread on average diagonal
Datapoints comparison with diagonal.

Fig. 6. Better or worse carrier loss (%)
Average placed on the zero line

A different representation of the above data places it around the control group (zero line) with gain values by carrier. Values are sorted and charted from left to right by increasing average loss. The chart depicts gains or worsening on a network based on the average control group’s performance – values are shown from left to right following better to worse loss values. The assumption of this analysis is that while a network’s conditions might be better or worse compared to other networks, the conditions tend to be equal across all carriers including the control group. While the carrier’s network is not the culprit causing additional loss, this analysis might be able to suggest whether those carriers peering with remote regions are deficient. Non-systemic issues with carriers will tend to cancel out with values being scattered equally above or below the zero line while systemic issues or gains will have a tendency to place a carrier consistently above or below it. The scatter plot highlights this assumption. More so, if we average gains or losses compared with the control group we expect the noise to cancel out.

Fig. 7. Average packet loss gains/losses by carrier November 2017
Averages determined for ALL datapoints or a cutoff at 4.5% control group applied.

Fig. 8. Average packet loss gains/losses by carrier October 2017
Averages determined for ALL datapoints or a cutoff at 4.5% control group applied.

For November 2017, based on ALL data-points, the best position in terms of loss was held by Centurylink, while other Tier 1s registered poorer results. Centurylink, after the cut off at the 4.5% control group level has been applied, kept the same position as in the ALL data-points representation. Stable results suggest that Centurylink has registered preponderantly lower value of loss for packets which transit it, below the 4.5%.

As in October, Centurylink is present in the winner group for both graphics. GTT has considerably worsened its results for ALL data-points and for the applied cut off at the 4.5% control group. Level 3 and Zayo improved their statistics for ALL data-points, but maintained negative tendency as in October.

Latency

For the Latency analysis we use a similar scatter plot to the one we used for Loss. It displays control group values on the diagonal while highlighting individual carrier measurements on the horizontal and on the vertical axis. Datapoints placed significantly and consistently below the average highlight better performing carriers while datapoints above the average highlight worse than average performance.

Fig. 9. Carrier latency with average group on the diagonal
Clusters of datapoints below diagonal highlight better performance

• NTT was mostly present with the latency diapazon of 88 – 250 ms
• Level 3 and Cogent have registered results which depict traffic from local to very long distances (65 – 280 ms);
• GTT has been mostly present within the latency diapazon of 125 – 200 ms;
• Telia is spotted within the latency diapazon of 60 – 180 ms;
• Centurylink has shown the latency between 85 -200 ms;
• Zayo and XO are spotted within the latency diapazon of 75 – 175 ms;
• Hurricane Electric has been mostly present within the 130 – 210 ms latency range.

The results above linger around the diagonal with the following observations:

Fig. 10. Average latency gains/losses by carrier Values averaged for the difference between carrier performance and the average group in that network.

The differences in latency from the control group shown above are averaged with the expectation that better or worse performance will cancel out if the differences are caused by measurement noise.

The results show that during the month of November in comparison with October:

• XO maintained the leading position, reducing its average RTT by ~11 ms. GTT, Telia and Level 3 improved their average latency for each packet by ~ 1,6 ms, ~ 1,5 ms and ~1, 4ms respectively;
• Cogent and NTT kept previous tendency, adding about 2 ms to their RTT;
• Hurricane Electric moved to the negative category adding ~ 4.3 ms to each packet.