commit 2d4b7ba817e6c398f5b0137d893d96957ac2adf7 Author: Mike Perry mikeperry-git@fscked.org Date: Sat Dec 3 18:13:37 2011 -0800
Clean up the spec a tad. --- NetworkScanners/BwAuthority/README.spec.txt | 15 ++++++++------- 1 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/NetworkScanners/BwAuthority/README.spec.txt b/NetworkScanners/BwAuthority/README.spec.txt index 5eb0661..a1b732e 100644 --- a/NetworkScanners/BwAuthority/README.spec.txt +++ b/NetworkScanners/BwAuthority/README.spec.txt @@ -371,10 +371,6 @@ The bandwidth authorities measure F_node: the filtered stream capacity through a given node (filtering is described in Section 1.6).
- In PID control, we add in this extra failure rate as a damper, to prevent - the PID control system from driving nodes to CPU overload. Once nodes - begin failing circuits, we want to stop devoting more capacity. - The PID Setpoint, or target for each node is F_avg: the average F_node value observed across the entire network.
@@ -412,9 +408,9 @@ hours. Since the bandwidth authorities also take on the order of hours to measure a slice of nodes, we do nothing special here.
- For Guard nodes however, clients have minimum rotation rate of 4-6 + For Guard nodes however, clients try to keep their Guards for 4-6 weeks. However, on the assumption that they rotate more frequently - than this, we set our Guard feedback interval to 2 weeks. + than this in practice, we set our Guard feedback interval to 2 weeks.
Guard measurements are also used without feedback whenever new measurements are available, to compensate for changes in Guard flag @@ -428,7 +424,7 @@ determining when to report a measurement.
The integral component, pid_error_sum is subjected to a decay factor - of 20% per interval, to prevent unbounded growth in cases without + per each interval, to prevent unbounded growth in cases without convergence.
The differential component is a simple delta, calculating by @@ -521,6 +517,11 @@ of the circ_error and the original pid_error as the new pid_error.
+ We use this to prevent the PID control system from driving nodes + to CPU overload. Once nodes begin failing circuits, we want to stop + devoting additional capacity to them (and decrease it proportional + to their failure rate relative to the rest of the network). + "bwauthbestratio=0" If absent, the larger of stream bandwidth vs filtered bandwidth is used to compute F_node.