User Guide

Ohno is a command line client, wrapper, and Python module for giving you formatted error messages. If you have a question, find a bug, or want to request a feature! This is an open source project and we are eager for your contribution. 🎉️

Usage

Once you have ohno installed (Installation) you can parse your errors fairly easily, either using ohno as a wrapper or post-run log parser.

Monitor

To monitor something that might error, you can do:

$ ohno monitor <command>

We provide a set of failing scripts you can run:

$ ohno monitor python examples/monitor/failure.py

The output generated (by default) is markdown, and the intention is that you can open a GitHub issue with versions, system information, and the error message.

### To Reproduce

```bash
$ python examples/monitor/failure.py
```

### Environment

- Version: Python 3.8.8
- Host:
  - architecture: x86_64
  - python-compiler: GCC 7.3.0
  - python-version: 3.8.8
  - cpu-count: 8
  - system: Linux
  - libc: glibc@2.31

### Errors

```bash
>> 1     error: We cannot do that thing.     2
```

Parse

You can also parse an existing log.

$ ohno parse examples/parse/spack-error.txt

### To Reproduce

```bash
$ <fill in command here>
```

### Environment


- Host:
  - architecture: x86_64
  - python-compiler: GCC 7.3.0
  - python-version: 3.8.8
  - cpu-count: 8
  - system: Linux
  - libc: glibc@2.31

### Errors

```bash
>> 2152     configure: error: no BPatch.h found; check path for Dyninst packa
>> 2154     configure: error: no tau_instrumentor found; check path for PDToo
>> 2178     configure: error: MPI Correctness Checking support cannot be buil
>> 2186     configure: error: no f2c.h found; check path for CLAPACK package
>> 2208     configure: error: no jvmti.h found; check path for JVMTI package
>> 2238     configure: error: no cuda.h found; check path for CUDA Toolkit fi
>> 2242     configure: error: no cuda_runtime_api.h found; check path for CUD
>> 2246     configure: error: no cupti.h found; check path for CUPTI package
>> 2256     configure: error: no ctool/ctool.h found; check path for CTool pa
>> 2293     configure: error: no bmi.h found; check path for BMI package firs
>> 2393     configure: error: no vtf3.h found; check path for VTF3 package fi
>> 14683    make[4]: *** [mpi_comm_spawn_multiple_f90.lo] Error 1
>> 14694    make[4]: *** [mpi_testall_f90.lo] Error 1
>> 14696    make[3]: *** [all-recursive] Error 1
>> 14698    make[2]: *** [all] Error 2
>> 14700    make[1]: *** [all-recursive] Error 1
>> 14702    make: *** [all-recursive] Error 1
>> 2152     configure: error: no BPatch.h found; check path for Dyninst packa
>> 2154     configure: error: no tau_instrumentor found; check path for PDToo
>> 2178     configure: error: MPI Correctness Checking support cannot be buil
>> 2186     configure: error: no f2c.h found; check path for CLAPACK package
>> 2208     configure: error: no jvmti.h found; check path for JVMTI package
>> 2238     configure: error: no cuda.h found; check path for CUDA Toolkit fi
>> 2242     configure: error: no cuda_runtime_api.h found; check path for CUD
>> 2246     configure: error: no cupti.h found; check path for CUPTI package
>> 2256     configure: error: no ctool/ctool.h found; check path for CTool pa
>> 2293     configure: error: no bmi.h found; check path for BMI package firs
>> 2393     configure: error: no vtf3.h found; check path for VTF3 package fi
>> 14686    make[4]: *** [mpi_testsome_f90.lo] Error 1
>> 14696    make[4]: *** [mpi_testall_f90.lo] Error 1
>> 14702    make[3]: *** [all-recursive] Error 1
>> 14704    make[2]: *** [all] Error 2
>> 14706    make[1]: *** [all-recursive] Error 1
>> 14708    make: *** [all-recursive] Error 1
```

The view you see above are the lines extracted that are determined to have meaningful content.

Python Usage

If you are working in Python, the same interactions can be done programatically. Here is how to monitor a command:

import ohno.main.client as client
monitor = client.Monitor()
res = monitor.run(['python', 'monitor/failure.py'])
print(res.parse())

You can explore the results object:

res.errors
res.warnings
res.to_dict()

And the original command is now a known task to the monitor:

monitor.tasks
{'python monitor/failure.py': <ohno.main.executor.shell.ShellExecutor at 0x7fd5541cce50>}
A task is parsed by a particular kind of result (output) you are interested in:

To do the same but for an error log already generated:

import ohno.main.client as client
monitor = client.Monitor()
res = monitor.load("parse/spack-error.txt")
print(res.parse())

This library is under development and we will have more documentation coming soon!