Mailing-List: contact user-help@thrift.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@thrift.apache.org
Received-SPF: pass (nike.apache.org: domain of jpaton2@gmail.com designates
 209.85.212.42 as permitted sender)
Content-Type: text/plain; charset=iso-8859-1
Mime-Version: 1.0 (Apple Message framework v1278)
Subject: Re: Python - Killing an active Thrift call
From: James Paton <jpaton2@gmail.com>
In-Reply-To: 
 <CAEExJ=yHA4q4kD4Zb=rPRe8wgaYe6uv-0akUpJe6ssWVRE6ZQg@mail.gmail.com>
Date: Thu, 14 Jun 2012 14:11:57 -0500
Content-Transfer-Encoding: quoted-printable
Message-Id: <C3FD4A7F-7900-444B-9E07-8FBD4ECF7406@gmail.com>
References: <4FDA30BB.1080500@subwest.com>
 <CAEExJ=yHA4q4kD4Zb=rPRe8wgaYe6uv-0akUpJe6ssWVRE6ZQg@mail.gmail.com>
To: user@thrift.apache.org

I asked a question like this awhile ago on this list. I was told that =
Thrift does not have a way to interrupt an ongoing RPC from the client. =
In the past I have done this by interrupting the thread that was making =
the call (this was in Java), but this is probably not quite right as I =
imagine it leaves the server spinning.

-- Jim

On Jun 14, 2012, at 2:06 PM, Dan Kuebrich wrote:

> I don't know that there's a built-in facility for this, but it sounds =
like
> what timeouts are often used for.  How are you receiving the info that =
the
> other side won't return? Can you use timeouts on the thrift client =
sockets
> for this?
>=20
> On Thu, Jun 14, 2012 at 2:43 PM, Steve <steve@subwest.com> wrote:
>=20
>> Hi,
>>=20
>> Is it possible to kill an active/open request in Python?
>>=20
>> I have locking code around a thrift call, but sometimes the other =
side
>> (server) doesn't return (not in my control).  I would like to force =
that
>> connection to close so that the thread can continue on.
>>=20
>> I've tried to close the transport, thinking it would fail out, but no =
luck.
>>=20
>> Thanks.
>>=20
>>=20