CLI tool to profile Terraform runs, written in Go.
Main features:
- Modern CLI (cobra-based) with autocomplete
- Read logs straight from your Terraform process (using pipe) or a log file
- Can generate global stats, resource-level stats or visualizations
- Provides many levels of granularity and aggregation and customizable outputs
Featured on: awesome-go | awesome-terraform
❱ brew tap datarootsio/tf-profile
❱ brew install tf-profile
❱ tf-profile --help
tf-profile is a CLI tool to profile Terraform runs
Usage:
tf-profile [command]- Head over to the releases page (https://github.com/QuintenBruynseraede/tf-profile/releases)
- Download the correct binary for your operating system
- Copy it to a path that is on your
$PATH. On a Linux system,/usr/local/binis the most common location.
If you want to try tf-profile without installing anything, you can run it using Docker (or similar).
❱ cat my_log_file.log | docker run -i qbruynseraede/tf-profile:0.2.0 stats
Key Value
Number of resources created 1510
Cumulative duration 36m19s
Longest apply time 7m18s
Longest apply resource time_sleep.foo[*]
...Optionally, define an alias:
❱ alias tf-profile=docker run -i qbruynseraede/tf-profile:0.2.0
❱ cat my_log_file.log | tf-profileThis requires at least version 1.20 of the go cli.
❱ git clone git@github.com:QuintenBruynseraede/tf-profile.git
❱ cd tf-profile && go build .
❱ sudo ln -s $(pwd)/tf-profile /usr/local/bin # Optional: only if you want to run tf-profile from other directories
❱ tf-profile --help
tf-profile is a CLI tool to profile Terraform runs
Usage:
tf-profile [command]tf-profile handles input from stdin and from files. These two commands are therefore equivalent:
❱ terraform apply -auto-approve | tf-profile table
❱ terraform apply -auto-approve > log.txt && tf-profile table log.txtFour major commands are supported:
- 🔗
tf-profile stats: provide general statistics about a Terraform run - 🔗
tf-profile table: provide detailed, resource-level statistics about a Terraform run - 🔗
tf-profile filter: filter logs to include only certain resources - 🔗
tf-profile graph: generate a visual overview of a Terraform run.
tf-profile stats is the most basic command. Given a Terraform log, it will only provide high-level statistics.
❱ terraform apply -auto-approve > log.txt
❱ tf-profile stats log.txt
Key Value
-----------------------------------------------------------------
Number of resources in configuration 1510
Cumulative duration 36m19s
Longest apply time 7m18s
Longest apply resource time_sleep.foo[*]
Resources marked for operation Create 892
Resources marked for operation None 18
Resources marked for operation Replace 412
Resources in state AllCreated 800
Resources in state Created 695
Resources in state Started 15
Resources in desired state 1492 out of 1510 (98.8%)
Resources not in desired state 18 out of 1510 (0.01%)
Number of top-level modules 13
Largest top-level module module.core[2]
Size of largest top-level module 170
Deepest module module.core[2].module.role[47]
Deepest module depth 2
Largest leaf module module.dbt[4]
Size of largest leaf module 40 For more information, refer to the reference for the stats command.
tf-profile table will parse a log and provide per-resource metrics.
❱ terraform apply -auto-approve > log.txt
❱ tf-profile table log.txt
resource n tot_time modify_started modify_ended desired_state operation final_state
aws_ssm_parameter.p6 1 0s 6 7 Created Replace Created
aws_ssm_parameter.p1 1 0s 7 5 Created Replace Created
aws_ssm_parameter.p3 1 0s 5 6 Created Replace Created
aws_ssm_parameter.p4 1 0s / 1 NotCreated Destroy NotCreated
aws_ssm_parameter.p5 1 0s 4 4 Created Modify Created
aws_ssm_parameter.p2 1 0s / / Created None Created For a full description of the options, see the reference page.
tf-profile filter filters logs to include only certain resources. Wildcards are supported to filter on multiple resources.
❱ tf-profile filter "module.*.null_resource.*" log.txt
# module.mod1.null_resource.foo will be created
+ resource "null_resource" "foo" {
...
}
# module.mod2.null_resource.bar will be created
+ resource "null_resource" "bar" {
...
}
module.mod1.null_resource.foo: Creating...
module.mod2.null_resource.bar: Creating...
module.mod1.null_resource.foo: Creation complete after 1s [id=foo]
module.mod2.null_resource.bar: Creation complete after 1s [id=bar]For a full description of the options, see the reference page.
tf-profile graph is used to visualize your terraform logs. It generates a Gantt-like chart that shows in which order resources were created. tf-profile does not actually create the final image, but generates a script file that Gnuplot understands.
❱ tf-profile graph my_log.log --out graph.png --size 2000,1000 | gnuplot
Disclaimer: Terraform's logs do not contain any absolute timestamps. We can only derive the order in which resources started and finished their modifications. Therefore, the output of tf-profile graph gives only a general indication of how long something actually took. In other words: the X axis is meaningless, apart from the fact that it's monotonically increasing.
