Authors: Paweł Pierścionek, Founder, Nodely and Simon Bonanno, Head of Business Intelligence, Algorand Foundation
The current KPI source
The Nodely analytics dataset is the source of the node count KPI, which is presented on Nodely's network dashboard and exposed via public API for others to consume.
The value is derived from unique node IP addresses that connect over 24-hour periods to Nodely-operated relays. The daily unique count is highly volatile, and a weekly max is used as the final value on the dashboard.
The theory behind the KPI
Nodes connect to 4 relays simultaneously, and one connection jumps randomly every 4 minutes. With a sufficient number of relays and a wide observation window, this allows for capturing all of the unique nodes.
The unique IP address count approximates the total node count, provided nodes do not change IP addresses and have long run times.
Sophisticated botnets utilizing short-lived node connections and a significant number of home-operated nodes with ephemeral IP addresses make the current counting method obsolete.
New node KPI
Each node counts: the validator node, the API node, and the bot node. KPI cannot discriminate between node types as each puts a load on the relay network or upcoming P2P mesh. We have to make sure that nodes are only represented proportionally to their run-time.
Let the KPI count full-time node equivalents, not unique nodes.
New node KPI, methodology
Decreasing the window time from one day to one hour is the simplest solution but poses many challenges - nodes have fewer opportunities to hit one of Nodely's relays. This makes the raw, unique IP address counter highly correlated with the number of relays operated by Nodely. Ultimately, the counter should be based on unified telemetry from all relays. Until that is addressed, the KPI will be using the Chao-1 formula for estimating animal populations with incomplete sampling data.
KPIs derived using this method should provide stable estimations with varying or small numbers of relays.
The daily mean of hourly node population size estimation will correspond to full-time node equivalents, representing fairly both long-lived and short-lived nodes.
New node KPI => 3700 node equivalents
Using a 7-day moving average to smooth out the final chart, the current estimate changes from 6700 unique nodes daily to around 3700 full-time node equivalents.
The difference between the previous value and the new value is primarily due to the frequently changing IP addresses of nodes in some countries as well as short-lived arbitrage bot connections that are now properly coalesced to full-time node equivalents.
The new KPI has the following benefits:
-
-
- Introduces new full-time node equivalents, does not invalidate the previous one
- Does not require Nodely to have telemetry access / run a lot of relays
- Should provide P2P node estimation with a few permissionless P2P Nodely relays
-
Disclaimer: The content provided in this newsletter is for informational purposes only. The information is provided by the Algorand Foundation and while we strive to keep the information up-to-date and correct, we make no representations or warranties of any kind, express or implied, about the completeness, accuracy, reliability, suitability, or availability with respect to the blog or the information, products, services, or related graphics contained in the newsletter for any purpose.
The content of this newsletter is not intended to be financial advice. You should not take any action before conducting your own research or consulting with a qualified professional. Any reliance you place on such information is therefore strictly at your own risk.
In no event will Algorand Foundation be liable for any loss or damage including without limitation, indirect, or consequential loss or damage, or any loss or damage whatsoever arising from loss of data or profits arising out of, or in connection with, the use of this newsletter.
Through this newsletter, you may be able to link to other websites which are not under the control of the Algorand Foundation. We have no control over the nature, content, and availability of those sites. The inclusion of any links does not imply a recommendation nor endorse the views expressed therein.