-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathCITATION.cff
36 lines (36 loc) · 1.58 KB
/
CITATION.cff
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
cff-version: 1.2.0
message: "If you use this software, please cite it as below."
authors:
- family-names: "Stendler"
given-names: "Maximilian"
title: "Tootbench"
date-released: 2023-03-13
url: "https://git.tu-berlin.de/stendler/tootbench"
preferred-citation:
type: report
authors:
- family-names: Stendler
given-names: Maximilian
title: "Tootbench - Report on Benchmarking the Overhead of Mastodon’s Federation"
url: "https://tubcloud.tu-berlin.de/s/qoxE9rcP32Qmntk"
year: 2023
month: 3
abstract: >
The rising popularity and growing user base of
the Fediverse and Mastodon in particular, begs the question
regarding the impact of federation on participating instances.
This report details the technical implementation, execution and
result evaluation of a benchmark of the overhead of Mastodons
federation. Mastodon was deployed in production mode using
Docker on multiple Compute Engine VMs on the Google Cloud
Platform.
Scenarios with varying number of instances and users are
evaluated regarding CPU utilization per instance, global CPU
utilization and latency of published messages distributed from
the server to the recipients. The global CPU utilization was the
same for a varying number of federating instances and the same
amount of global users. But individual instances can be relieved
when users are distributed among them. A significant 4 to 10
times higher latency of message distribution was measured due
to federation, which is still deemed to be acceptable for the human
users of this kind of application.