public inbox for kvm@vger.kernel.org
 help / color / mirror / Atom feed
From: Lucas Meneghel Rodrigues <lmr@redhat.com>
To: autotest@test.kernel.org
Cc: kvm@vger.kernel.org, Michael Goldish <mgoldish@redhat.com>
Subject: Re: [KVM-AUTOTEST PATCH 01/06] kvm_vm.py: create a lock file to avoid VM creation in parallel
Date: Wed, 10 Jun 2009 16:29:05 -0300	[thread overview]
Message-ID: <1244662145.22524.11.camel@localhost.localdomain> (raw)
In-Reply-To: <1244507691-9575-1-git-send-email-lmr@redhat.com>

On Mon, 2009-06-08 at 21:34 -0300, Lucas Meneghel Rodrigues wrote:
> VM.create() does a few things (such as finding free ports) which
> are not safe to execute in parallel. Use a lock file to make
> sure this doesn't happen. The lock is released only after the
> VM is started or fails to start.

Applied, thanks!

> Signed-off-by: Michael Goldish <mgoldish@redhat.com>
> ---
>  client/tests/kvm/kvm_vm.py |   91 ++++++++++++++++++++++++--------------------
>  1 files changed, 50 insertions(+), 41 deletions(-)
> 
> diff --git a/client/tests/kvm/kvm_vm.py b/client/tests/kvm/kvm_vm.py
> index eb9717b..e2b684c 100644
> --- a/client/tests/kvm/kvm_vm.py
> +++ b/client/tests/kvm/kvm_vm.py
> @@ -335,51 +335,60 @@ class VM:
>                      logging.error("Actual MD5 sum differs from expected one")
>                      return False
>  
> -        # Handle port redirections
> -        redir_names = kvm_utils.get_sub_dict_names(params, "redirs")
> -        host_ports = kvm_utils.find_free_ports(5000, 6000, len(redir_names))
> -        self.redirs = {}
> -        for i in range(len(redir_names)):
> -            redir_params = kvm_utils.get_sub_dict(params, redir_names[i])
> -            guest_port = int(redir_params.get("guest_port"))
> -            self.redirs[guest_port] = host_ports[i]
> -
> -        # Find available VNC port, if needed
> -        if params.get("display") == "vnc":
> -            self.vnc_port = kvm_utils.find_free_port(5900, 6000)
> -
> -        # Make qemu command
> -        qemu_command = self.make_qemu_command()
> +        # Make sure the following code is not executed by more than one thread
> +        # at the same time
> +        lockfile = open("/tmp/kvm-autotest-vm-create.lock", "w+")
> +        fcntl.lockf(lockfile, fcntl.LOCK_EX)
>  
> -        # Is this VM supposed to accept incoming migrations?
> -        if for_migration:
> -            # Find available migration port
> -            self.migration_port = kvm_utils.find_free_port(5200, 6000)
> -            # Add -incoming option to the qemu command
> -            qemu_command += " -incoming tcp:0:%d" % self.migration_port
> -
> -        logging.debug("Running qemu command:\n%s" % qemu_command)
> -        (status, pid, output) = kvm_utils.run_bg(qemu_command, None,
> -                                                 logging.debug, "(qemu) ")
> -
> -        if status:
> -            logging.debug("qemu exited with status %d" % status)
> -            logging.error("VM could not be created -- qemu command"
> -                          " failed:\n%s" % qemu_command)
> -            return False
> +        try:
> +            # Handle port redirections
> +            redir_names = kvm_utils.get_sub_dict_names(params, "redirs")
> +            host_ports = kvm_utils.find_free_ports(5000, 6000, len(redir_names))
> +            self.redirs = {}
> +            for i in range(len(redir_names)):
> +                redir_params = kvm_utils.get_sub_dict(params, redir_names[i])
> +                guest_port = int(redir_params.get("guest_port"))
> +                self.redirs[guest_port] = host_ports[i]
> +
> +            # Find available VNC port, if needed
> +            if params.get("display") == "vnc":
> +                self.vnc_port = kvm_utils.find_free_port(5900, 6000)
> +
> +            # Make qemu command
> +            qemu_command = self.make_qemu_command()
> +
> +            # Is this VM supposed to accept incoming migrations?
> +            if for_migration:
> +                # Find available migration port
> +                self.migration_port = kvm_utils.find_free_port(5200, 6000)
> +                # Add -incoming option to the qemu command
> +                qemu_command += " -incoming tcp:0:%d" % self.migration_port
> +
> +            logging.debug("Running qemu command:\n%s", qemu_command)
> +            (status, pid, output) = kvm_utils.run_bg(qemu_command, None,
> +                                                     logging.debug, "(qemu) ")
> +
> +            if status:
> +                logging.debug("qemu exited with status %d", status)
> +                logging.error("VM could not be created -- qemu command"
> +                              " failed:\n%s", qemu_command)
> +                return False
>  
> -        self.pid = pid
> +            self.pid = pid
>  
> -        if not kvm_utils.wait_for(self.is_alive, timeout, 0, 1):
> -            logging.debug("VM is not alive for some reason")
> -            logging.error("VM could not be created with"
> -                          " command:\n%s" % qemu_command)
> -            self.destroy()
> -            return False
> +            if not kvm_utils.wait_for(self.is_alive, timeout, 0, 1):
> +                logging.debug("VM is not alive for some reason")
> +                logging.error("VM could not be created with"
> +                              " command:\n%s", qemu_command)
> +                self.destroy()
> +                return False
>  
> -        logging.debug("VM appears to be alive with PID %d" % self.pid)
> +            logging.debug("VM appears to be alive with PID %d", self.pid)
> +            return True
>  
> -        return True
> +        finally:
> +            fcntl.lockf(lockfile, fcntl.LOCK_UN)
> +            lockfile.close()
>  
> 
>      def send_monitor_cmd(self, command, block=True, timeout=20.0):
-- 
Lucas Meneghel Rodrigues
Software Engineer (QE)
Red Hat - Emerging Technologies


      parent reply	other threads:[~2009-06-10 19:29 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-09  0:34 [KVM-AUTOTEST PATCH 01/06] kvm_vm.py: create a lock file to avoid VM creation in parallel Lucas Meneghel Rodrigues
2009-06-09  0:34 ` [KVM-AUTOTEST PATCH 02/06] kvm_runtest_2.py: Use env filename specified by the 'env' parameter Lucas Meneghel Rodrigues
2009-06-09  0:34   ` [KVM-AUTOTEST PATCH 03/06] kvm_vm.py: add new VM parameter 'x11_display' that controls $DISPLAY Lucas Meneghel Rodrigues
2009-06-09  0:34     ` [KVM-AUTOTEST PATCH 04/06] VM.create(): always destroy() the VM before attempting to start it Lucas Meneghel Rodrigues
2009-06-09  0:34       ` [KVM-AUTOTEST PATCH 05/06] kvm_utils.py: remote_login(): improve regular expression matching Lucas Meneghel Rodrigues
2009-06-09  0:34         ` [KVM-AUTOTEST PATCH 06/06] kvm_config: Make split_and_strip function to strip the right quotes Lucas Meneghel Rodrigues
2009-06-10 19:30           ` Lucas Meneghel Rodrigues
2009-06-10 19:30         ` [KVM-AUTOTEST PATCH 05/06] kvm_utils.py: remote_login(): improve regular expression matching Lucas Meneghel Rodrigues
2009-06-10 19:29       ` [KVM-AUTOTEST PATCH 04/06] VM.create(): always destroy() the VM before attempting to start it Lucas Meneghel Rodrigues
2009-06-10 19:29     ` [KVM-AUTOTEST PATCH 03/06] kvm_vm.py: add new VM parameter 'x11_display' that controls $DISPLAY Lucas Meneghel Rodrigues
2009-06-10 19:29   ` [KVM-AUTOTEST PATCH 02/06] kvm_runtest_2.py: Use env filename specified by the 'env' parameter Lucas Meneghel Rodrigues
2009-06-10 19:29 ` Lucas Meneghel Rodrigues [this message]

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=1244662145.22524.11.camel@localhost.localdomain \
    --to=lmr@redhat.com \
    --cc=autotest@test.kernel.org \
    --cc=kvm@vger.kernel.org \
    --cc=mgoldish@redhat.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox