lttng-dev.lists.lttng.org archive mirror
 help / color / mirror / Atom feed
From: Philippe Proulx via lttng-dev <lttng-dev@lists.lttng.org>
To: Dave Bort <dbort@dbort.com>
Cc: lttng-dev <lttng-dev@lists.lttng.org>
Subject: Re: [lttng-dev] babeltrace2: reading metadata + events from a single CTF file
Date: Tue, 8 Sep 2020 10:53:57 -0400	[thread overview]
Message-ID: <CAB4xu_32eMKy2sndQNtHmVZ4go3rs2Wu-hBv09xjTYu_U8fPGg@mail.gmail.com> (raw)
In-Reply-To: <CAM-pDOyr2VN94eWJahkRX47rTp=YSOwvyZL5iVO2bETaaLx+1Q@mail.gmail.com>

On Mon, Sep 7, 2020 at 8:35 AM Dave Bort via lttng-dev
<lttng-dev@lists.lttng.org> wrote:
>
> Hello! I'm adding CTF-compatible tracing to a custom/experimental OS and have a question:
>
> Can babeltrace2 read from a single CTF output file that contains both the metadata stream and the binary trace streams?

No.

>
> My little trace system would like to bundle everything related to a trace into a single file, making it easier to pass around (since there's not much filesystem/network support yet). The CTF spec provides a way to embed the metadata stream in the same bitstream as the data streams, but https://babeltrace.org/docs/v2.0/man7/babeltrace2-source.ctf.fs.7/ says that the `metadata` file needs to be separate. I see that LTTng manages trace output as a collection of files, with a separate `metadata` file.
>
> I'd rather avoid writing a custom tool to find and extract the metadata stream from the full stream before passing the data on to babeltrace2 or TraceCompass, but so far it looks like I may need to.

You will need to.

>
> Would it be appropriate to add support for this to babeltrace2-source.ctf.fs? Or would this warrant a new source.ctf plugin? If I'm going to write some new code to handle this situation, I may as well try to add it to the project.

This would need to be specified at the CTF level.

Storing a CTF trace on the file system as a single file has been
requested a few times, so we _might_ come up with a specified way to do
it as an addendum to CTF 2 (which is on the roadmap, but not
finalized/released yet).

As of CTF 1.8, the de facto standard way to store a trace on the file
system is to write its metadata stream to the `metadata` file and each
individual data stream to its own file of which the name doesn't begin
with `.`. This is also explained in the "Input" section of
`babeltrace2-source.ctf.fs(7)` [1].

>
> Thank you! I'm really looking forward to benefiting from the CTF/babeltrace/LTTng/TraceCompass work you've all done over the years. :)

I can assure you that the EfficiOS team is very glad to hear that.

Phil

[1]: https://babeltrace.org/docs/v2.0/man7/babeltrace2-source.ctf.fs.7/#doc-input

>
> --Dave Bort
> _______________________________________________
> lttng-dev mailing list
> lttng-dev@lists.lttng.org
> https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev
_______________________________________________
lttng-dev mailing list
lttng-dev@lists.lttng.org
https://lists.lttng.org/cgi-bin/mailman/listinfo/lttng-dev

  reply	other threads:[~2020-09-08 14:54 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-09-06 20:46 Dave Bort via lttng-dev
2020-09-08 14:53 ` Philippe Proulx via lttng-dev [this message]
2020-09-08 20:40   ` Dave Bort via lttng-dev

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=CAB4xu_32eMKy2sndQNtHmVZ4go3rs2Wu-hBv09xjTYu_U8fPGg@mail.gmail.com \
    --to=lttng-dev@lists.lttng.org \
    --cc=dbort@dbort.com \
    --cc=eeppeliteloop@gmail.com \
    --subject='Re: [lttng-dev] babeltrace2: reading metadata + events from a single CTF file' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).