App Install
Measure app installation speed. This scenario has two roles:
minimal: Installs a simple app which implements initialisation callbacks but otherwise doesn't contain a lot of code.large: Installs a larger app that contains some dependencies and some generated data as padding to make the bundle larger.
These are intended to catch a compilation slowdown or issues with copying around large WASMs (e.g. accidental cloning or tracing). In each case, the behaviour will uninstall the app it installed so that it can re-install it on the next iteration.
-
minimal(1 agent)
App Install
Measure app installation speed. This scenario has two roles:
minimal: Installs a simple app which implements initialisation callbacks but otherwise doesn't contain a lot of code.large: Installs a larger app that contains some dependencies and some generated data as padding to make the bundle larger.
These are intended to catch a compilation slowdown or issues with copying around large WASMs (e.g. accidental cloning or tracing). In each case, the behaviour will uninstall the app it installed so that it can re-install it on the next iteration.
-
large(1 agent)
DHT Sync Lag
Measure lag time between an agent publishing data and other peers being able to see it. This scenario has two roles:
write: A simple job that just creates entries with a timestamp field. Those entries are linked to a known base hash. For each write, the metricws.custom.dht_sync_sent_countis incremented.record_lag: A job that repeatedly queries for links from the known base hash. It keeps track of records that it has seen and when a new record is found, and calculates the time difference between the timestamp of the new record and the current time. That time difference is then recorded as a custom metric calledwt.custom.dht_sync_lag.
After each behaviour loop the metric ws.custom.dht_sync_recv_count is incremented.
-
record_lag(5 agents) -
write(5 agents)
agent:uhCAkWi-t8tNx5GEhOTzSMNhoVR1LPXllFl6shjH6CVbYK46ZSL9v
agent:uhCAkbGCnezK-FAtOglxstZlo_OeyixPaqbow5qHZdo4hB0KXa_rL
agent:uhCAkgGdm5794bjIaiuQopJv-aOckM5uDFRokDftNPid_ErkLbrcf
agent:uhCAko2IGJ6Bd-EUITzQf1-IctWv3roJiK8JyQI6a25p3q5M4TJ9I
agent:uhCAkzNYCcAxLCHq4fZ-wQ8gzgzmAoFpA7KDgMR7qnFuA6TpnWSyQ
agent:uhCAk71AxgB7arLxM0TnjmFrEt_TfGd21yrEU3MfOCpVltCzeWwHM
agent:uhCAkLctbHi-SJXjRlMnnjDD_3idFbQXUOBM_JxzuLfmQjlIrzeba
agent:uhCAkVBVl2LZ5ZmXWbQZorVugardqAb1H6UX-jcjIftNRdSR_uuP6
agent:uhCAkn-LoqDNurFObTvT-kOqEo5E1PJm9YbQISadE1yE_5QiL5YNG
agent:uhCAkuYZzcBvvs9w5EXIKb_wColTsjUYQAUMN5KNOi-obULQOH_lY
agent:uhCAk71AxgB7arLxM0TnjmFrEt_TfGd21yrEU3MfOCpVltCzeWwHM
agent:uhCAkLctbHi-SJXjRlMnnjDD_3idFbQXUOBM_JxzuLfmQjlIrzeba
agent:uhCAkVBVl2LZ5ZmXWbQZorVugardqAb1H6UX-jcjIftNRdSR_uuP6
agent:uhCAkn-LoqDNurFObTvT-kOqEo5E1PJm9YbQISadE1yE_5QiL5YNG
agent:uhCAkuYZzcBvvs9w5EXIKb_wColTsjUYQAUMN5KNOi-obULQOH_lY
First Call
This scenario installs a simple app which implements the init callback, and measures the time elapsed between start of install and calling init. It will uninstall the app it installed so that it can re-install it on the next iteration. This is required to re-run the initialisation callback.
- Default (1 agent)
init.Local Signals
This scenario is for testing the speed and reliability of local signals. It uses a zome with a function that will emit 10,000 signals. The scenario measures how long it takes to send the signals and how many have been received by the time the zome call ends.
- Default (1 agent)
No scenario template for mixed_arc_get_agent_activity!
Remote Call Rate
Test the throughput of remote_call operations. Each agent in this scenario waits for a certain number of peers to be available or for up to two minutes, whichever happens first, before starting its behaviour.
- Default (1 agent)
agent:uhCAkfDOU61JiLpVYMTaR3um9GsaJGuQYtHlfGlgQXW8X2gCl12yb
agent:uhCAkhZ8EEy86R3vjNOtKMuKrdJSsBZCJi6tIjCJGzBqtJaBUmkir
agent:uhCAkfDOU61JiLpVYMTaR3um9GsaJGuQYtHlfGlgQXW8X2gCl12yb
agent:uhCAkhZ8EEy86R3vjNOtKMuKrdJSsBZCJi6tIjCJGzBqtJaBUmkir
Single-write, many-read
Creates an entry in the agent setup, then read one's own record back. This tests the maximum read performance for reading back local data when the chain is short.
- Default (1 agent)
read call timingread call rate (10s)Write/get_agent_activity
A scenario where write peers write entries, while get_agent_activity peers each query a single write agent's activity with get_agent_activity.
Before a target write peer and the requesting get_agent_activity peer are in sync, this will measure the get_agent_activity call performance over a network. Once a write peer reaches sync with a get_agent_activity peer, the write peer will publish their actions and entries, and so the get_agent_activity calls will likely have most of the data they need locally. At that point this measures the database query performance and code paths through host functions.
-
get_agent_activity(1 agent) -
write(1 agent)
action_seqwrite peer's source chain.get_agent_activity_full zome call timingwrite peer's source chain.agent:uhCAkSSHyQ5nScpIH4C3BactojdqVjCTnIXWJ8MMSfJs9_lEzdqJU
Write/Query
Creates an entry, then queries the source chain and performs a simple operation on the entries, then repeats.
- Default (1 agent)
Write/Read
Creates an entry and reads it back, then repeats.
- Default (1 agent)
Write validated
Creates an entry, updates it and links the create to the update, then repeats. Each of the three actions is validated by the included hApp.
- Default (1 agent)
Write Validated must_get_agent_activity
A scenario where write agents create entries in batches of 10, while must_get_agent_activity agents each pick a random write agent and repeatedly attempt to create an entry that references the chain top of their latest batch. This reference means that the entry's validation function needs to make a must_get_agent_activity call.
The purpose of this scenario is to measure the time it takes for published agent activity data to be gossiped among authorities and become available to peers that query it via must_get_agent_activity.
This test is similar to Mixed-Arc must_get_agent_activity, but all agents are full-arc.
-
must_get_agent_activity(1 agent) -
write(1 agent)
action_seqaction_seq of a write agent successfully retrieved by a must_get_agent_activity agent. This reflects the DHT's ability to propagate agent activity ops and make them available to querying peers.write agent's creation of a batch and a must_get_agent_activity agent's successful discovery of the batch and creation/self-validation of a new entry that depends on it.must_get_agent_activity_agent:uhCAkM73YJMA0mD9rA8DKvMzQO6C3M4Qfb06DpA60staIrJdI5zxp
must_get_agent_activity agent was able to discover batches and create/self-validate new entries that depend on them.must_get_agent_activity_agent:uhCAkM73YJMA0mD9rA8DKvMzQO6C3M4Qfb06DpA60staIrJdI5zxp
create_validated_sample_entry zome call timingwrite agent's source chain.agent:uhCAkM73YJMA0mD9rA8DKvMzQO6C3M4Qfb06DpA60staIrJdI5zxp
must_get_agent_activity calls.agent:uhCAkM73YJMA0mD9rA8DKvMzQO6C3M4Qfb06DpA60staIrJdI5zxp
Zero-Arc Create Data
zero: A zero-arc conductor that just creates entries with a timestamp field. Those entries are linked to a known base hash so that full-arc nodes can retrieve them.full: A full-arc conductor that reads the entries created by the zero-arc node(s) and records the time lag between when the entry had been created and when it was first discovered.
-
full(4 agents) -
zero(6 agents)
agent:uhCAkJXlpyqPvax9DvLWnwF0PQz2ifFunix-H0pV3LekOlIBE49DK
agent:uhCAkT6xSMKv3Bm7Bx-mZJOgizNj6XvErmQ_lSOkMtdD972sOt8Pg
agent:uhCAkcrsKP_YvnfTL8zzVC-2srPKY09IaADO5OLk78WWFoh6x3qUa
agent:uhCAkf0Msfgjwpx6iPFY0TUQqemq3b6s0qBbsZMAxxzUkxTkVPB-C
agent:uhCAko0_nS8c0wE0-P4pdi5VGEH2m7-RmDzFSr8IdQbZn38Mn7PLd
agent:uhCAkxj1TGLjVF92gqmsghKVxGEBumCFdGcRJukkd2MVA_9Z6tgpA
get_timed_local_entries zome function.agent:uhCAk3xrQ6fCP8_LnCW9XEhzCLUwhQWjzWwKpr5Wo-vv6t8jw81Bd
agent:uhCAkKAwvPGUQmGlpVoYAayUtxr0ZvDVJsMDpKzhv3em7Az37_j_b
agent:uhCAkU_scCJj7M5kbGUjwN1gMvWz7OQO6DriyYcuMQ75MwwVd7nvC
agent:uhCAktyYzKh9CVienX3Z_xhaUo-zlLuhCV_XiGVmvyRRVooJ45Q-t
agent:uhCAk3xrQ6fCP8_LnCW9XEhzCLUwhQWjzWwKpr5Wo-vv6t8jw81Bd
agent:uhCAkKAwvPGUQmGlpVoYAayUtxr0ZvDVJsMDpKzhv3em7Az37_j_b
agent:uhCAkU_scCJj7M5kbGUjwN1gMvWz7OQO6DriyYcuMQ75MwwVd7nvC
agent:uhCAktyYzKh9CVienX3Z_xhaUo-zlLuhCV_XiGVmvyRRVooJ45Q-t
Zero-arc Create Data (Validated)
A zero-arc/full-arc mixed scenario where zero-arc nodes create data that gets validated and full-arc nodes read the data. The scenario has two roles:
zero: A zero-arc conductor that just creates entries with a timestamp field. Those entries are linked to a known base hash so that full-arc nodes can retrieve them.full: A full-arc conductor that reads the entries created by the zero-arc node(s) and records the time lag between when the entry had been created and when it was first discovered.
-
full(4 agents) -
zero(6 agents)
agent:uhCAk6DQzt0bBZSQjFSNPqunpx9UtXd_Z6H2sMo9bfw6Ldpo5sITC
agent:uhCAkB23i9UVJlqAkEX5WLllbPM0WSrLhTn0YWxY0U2nFkAXgfBTs
agent:uhCAkObGpracEEupG-WpPWGG1e2jmjnNJ784uUycTtjjOTHWSosGN
agent:uhCAklqqTXkYSTgrrrH0GhNsjHUmqC1g1iPA_Txp5qUdM6lPUONok
agent:uhCAkp-zSjgwOkksoX12Sa32PFUewrvycSh9W3Qz2uvDrc7N2kLu-
get_timed_local_entries zome function.agent:uhCAk8QiScV6n-s85_ZntR0fbEQ_Tede2tBQFOyqZ0FrMVgRv-coV
agent:uhCAkEGDsitfv6xt31a8u_dOXiZxapqaaRBxFBFr0Hs0Z30s4wznp
agent:uhCAkgUyj4-wEYOzmvocAK5s0-k_jIv4BzX-1T-uHZ-TCMoK1ihhG
agent:uhCAk8QiScV6n-s85_ZntR0fbEQ_Tede2tBQFOyqZ0FrMVgRv-coV
agent:uhCAkEGDsitfv6xt31a8u_dOXiZxapqaaRBxFBFr0Hs0Z30s4wznp
agent:uhCAkgUyj4-wEYOzmvocAK5s0-k_jIv4BzX-1T-uHZ-TCMoK1ihhG
Zome Call, Single Value
Calls a zome function that returns a fixed value. This tests the maximum performance of zome calls when the zome function is not calling into Holochain other otherwise doing any work.
- Default (1 agent)