4.2 KiB
| layout | page_title | sidebar_current | description |
|---|---|---|---|
| docs | Debugging | docs-internals-debug | Terraform has detailed logs which can be enabled by setting the TF_LOG environment variable to any value. This will cause detailed logs to appear on stderr |
Debugging Terraform
Terraform has detailed logs which can be enabled by setting the TF_LOG environment variable to any value. This will cause detailed logs to appear on stderr.
You can set TF_LOG to one of the log levels TRACE, DEBUG, INFO, WARN or ERROR to change the verbosity of the logs.
Setting TF_LOG to JSON outputs logs at the TRACE level or higher, and uses a parseable JSON encoding as the formatting.
~> Warning: The JSON encoding of log files is not considered a stable interface. It may change at any time, without warning. It is meant to support tooling that will be forthcoming, and that tooling is the only supported way to interact with JSON formatted logs.
Logging can be enabled separately for terraform itself and the provider plugins
using the TF_LOG_CORE or TF_LOG_PROVIDER environment variables. These take
the same level arguments as TF_LOG, but only activate a subset of the logs.
To persist logged output you can set TF_LOG_PATH in order to force the log to always be appended to a specific file when logging is enabled. Note that even when TF_LOG_PATH is set, TF_LOG must be set in order for any logging to be enabled.
If you find a bug with Terraform, please include the detailed log by using a service such as gist.
Interpreting a Crash Log
If Terraform ever crashes (a "panic" in the Go runtime), it saves a log file
with the debug logs from the session as well as the panic message and backtrace
to crash.log. Generally speaking, this log file is meant to be passed along
to the developers via a GitHub Issue. As a user, you're not required to dig
into this file.
However, if you are interested in figuring out what might have gone wrong before filing an issue, here are the basic details of how to read a crash log.
The most interesting part of a crash log is the panic message itself and the
backtrace immediately following. So the first thing to do is to search the file
for panic: , which should jump you right to this message. It will look
something like this:
panic: runtime error: invalid memory address or nil pointer dereference
goroutine 123 [running]:
panic(0xabc100, 0xd93000a0a0)
/opt/go/src/runtime/panic.go:464 +0x3e6
github.com/hashicorp/terraform/builtin/providers/aws.resourceAwsSomeResourceCreate(...)
/opt/gopath/src/github.com/hashicorp/terraform/builtin/providers/aws/resource_aws_some_resource.go:123 +0x123
github.com/hashicorp/terraform/helper/schema.(*Resource).Refresh(...)
/opt/gopath/src/github.com/hashicorp/terraform/helper/schema/resource.go:209 +0x123
github.com/hashicorp/terraform/helper/schema.(*Provider).Refresh(...)
/opt/gopath/src/github.com/hashicorp/terraform/helper/schema/provider.go:187 +0x123
github.com/hashicorp/terraform/rpc.(*ResourceProviderServer).Refresh(...)
/opt/gopath/src/github.com/hashicorp/terraform/rpc/resource_provider.go:345 +0x6a
reflect.Value.call(...)
/opt/go/src/reflect/value.go:435 +0x120d
reflect.Value.Call(...)
/opt/go/src/reflect/value.go:303 +0xb1
net/rpc.(*service).call(...)
/opt/go/src/net/rpc/server.go:383 +0x1c2
created by net/rpc.(*Server).ServeCodec
/opt/go/src/net/rpc/server.go:477 +0x49d
The key part of this message is the first two lines that involve hashicorp/terraform. In this example:
github.com/hashicorp/terraform/builtin/providers/aws.resourceAwsSomeResourceCreate(...)
/opt/gopath/src/github.com/hashicorp/terraform/builtin/providers/aws/resource_aws_some_resource.go:123 +0x123
The first line tells us that the method that failed is
resourceAwsSomeResourceCreate, which we can deduce that involves the creation
of a (fictional) aws_some_resource.
The second line points to the exact line of code that caused the panic, which--combined with the panic message itself--is normally enough for a developer to quickly figure out the cause of the issue.
As a user, this information can help work around the problem in a pinch, since it should hopefully point to the area of the code base in which the crash is happening.