[erlang-questions] Etop issues: Erlang top got garbage

Scott Lystig Fritchie <>
Thu Sep 18 22:46:08 CEST 2008

Brian Troutwine <> wrote:

bt> I'm attempting to use etop to debug a memory issue I am having, but
bt> the output of etop is almost impossible to read. I get many, many
bt> messages of this sort:

bt> Erlang top got garbage {trace_ts,<4549.52.0>,out,
bt> {gen_server,loop,6}, {1221,89025,387602}}

Yes, I've noticed that too.  The OTP folks almost certainly have more
insight than I do about the cause.

I redirect to a file and/or use "grep -v" when this happens.  Easier to
read that way.  :-)

I only see this when the system being examined is exceptionally busy.
I've assumed it's because it's so busy that trace messages are being
lost (or intentionally dropped) somewhere.  But that's just a guess.
Using "etop" on very busy servers is the norm, so it's more difficult to
give my usual reply to the "Doctor, it hurts when..." problem.

If push comes to shove, edit out the source code for the etop app and
remove the message?


More information about the erlang-questions mailing list