v2: The script is now a little bit safer so it doesn't conflicts with
network daemon trying to set configurations at the same time.
This patch fixes the behavior of the hv_set_ifconfig script when setting
the interface ip. Sometimes the interface has already been configured by
network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
exists error"; in order to avoid this error this patch makes sure double
checks the interface before trying anything.
Signed-off-by: Eduardo Otubo <[email protected]>
---
tools/hv/hv_set_ifconfig.sh | 44 ++++++++++++++++++++++++++++++++++++--------
1 file changed, 36 insertions(+), 8 deletions(-)
diff --git a/tools/hv/hv_set_ifconfig.sh b/tools/hv/hv_set_ifconfig.sh
index 735aafd64a3f..3a04b91f61e3 100755
--- a/tools/hv/hv_set_ifconfig.sh
+++ b/tools/hv/hv_set_ifconfig.sh
@@ -46,19 +46,47 @@
# is expected to return the configuration that is set via the SET
# call.
#
+interface=$(echo $1 | awk -F - '{ print $2 }')
+current_ip=$(ip addr show $interface|grep "inet ");
+config_file_ip=$(grep IPADDR $1|cut -d"=" -f2);
+current_ipv6=$(ip addr show $interface|grep "inet6 ");
+config_file_ipv6=$(grep IPV6ADDR $1|cut -d"=" -f2);
+config_file_ipv6_netmask=$(grep IPV6NETMASK $1|cut -d"=" -f2);
+config_file_ipv6=${config_file_ipv6}/${config_file_ipv6_netmask};
-echo "IPV6INIT=yes" >> $1
-echo "NM_CONTROLLED=no" >> $1
-echo "PEERDNS=yes" >> $1
-echo "ONBOOT=yes" >> $1
+network_service_state=$(/bin/systemctl is-active network);
+while [[ ${network_service_state} == "activating" \
+ || ${network_service_state} == "deactivating" ]]; do
+ # Network script is still working. let's wait a bit.
+ # The default timeout for systemd is 90s.
+ sleep 30s;
+ ((i++));
+ network_service_state=$(/bin/systemctl is-active network);
-cp $1 /etc/sysconfig/network-scripts/
+ # If network service doens't come up or down in 90s we log the
+ # error and give up.
+ if [[ $i == 3 ]]; then
+ logger "Couldn't set IP address for fail-over interface"\
+ " because network daemon might be busy. Try to"\
+ " if-down $interface && if-up $interface"\
+ " manually later.";
+ exit 1;
+ fi
+done
+# Only set the IP if it's not configured yet.
+if [[ $(test "${current_ip#*$config_file_ip}") == "$config_file_ip" \
+ || $(test "${current_ipv6#*$config_file_ipv6}") == "$current_ipv6" ]]; then
+ echo "IPV6INIT=yes" >> $1
+ echo "NM_CONTROLLED=no" >> $1
+ echo "PEERDNS=yes" >> $1
+ echo "ONBOOT=yes" >> $1
-interface=$(echo $1 | awk -F - '{ print $2 }')
+ cp $1 /etc/sysconfig/network-scripts/
-/sbin/ifdown $interface 2>/dev/null
-/sbin/ifup $interface 2>/dev/null
+ /sbin/ifdown $interface 2>/dev/null
+ /sbin/ifup $interface 2>/dev/null
+fi
--
2.13.5
On Mon, Aug 28, Eduardo Otubo wrote:
> + sleep 30s;
Was this runtime tested?
Once this sleep(1) is done, HV_UTIL_TIMEOUT kicks in and the daemon dies.
Olaf
On Mon, 28 Aug 2017 12:01:21 +0200
Eduardo Otubo <[email protected]> wrote:
> v2: The script is now a little bit safer so it doesn't conflicts with
> network daemon trying to set configurations at the same time.
>
> This patch fixes the behavior of the hv_set_ifconfig script when setting
> the interface ip. Sometimes the interface has already been configured by
> network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> exists error"; in order to avoid this error this patch makes sure double
> checks the interface before trying anything.
>
> Signed-off-by: Eduardo Otubo <[email protected]>
Adding new dependency on systemd is not going to make this script
even less useful. I wonder why the script still exists at all? Most of the
Linux distro's can already setup HV networking without it.
> -----Original Message-----
> From: Stephen Hemminger [mailto:[email protected]]
> Sent: Monday, August 28, 2017 11:16 AM
> To: Eduardo Otubo <[email protected]>
> Cc: [email protected]; [email protected]; Haiyang
> Zhang <[email protected]>; Stephen Hemminger
> <[email protected]>; David Miller <[email protected]>
> Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
>
> On Mon, 28 Aug 2017 12:01:21 +0200
> Eduardo Otubo <[email protected]> wrote:
>
> > v2: The script is now a little bit safer so it doesn't conflicts with
> > network daemon trying to set configurations at the same time.
> >
> > This patch fixes the behavior of the hv_set_ifconfig script when
> setting
> > the interface ip. Sometimes the interface has already been configured
> by
> > network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> > exists error"; in order to avoid this error this patch makes sure
> double
> > checks the interface before trying anything.
> >
> > Signed-off-by: Eduardo Otubo <[email protected]>
>
> Adding new dependency on systemd is not going to make this script
> even less useful. I wonder why the script still exists at all? Most of
> the
> Linux distro's can already setup HV networking without it.
>
This script is used by a host to inject IP into guests. KY knows more
details about it.
On Mon, Aug 28, 2017 at 03:56:44PM +0000, Haiyang Zhang wrote:
>
>
> > -----Original Message-----
> > From: Stephen Hemminger [mailto:[email protected]]
> > Sent: Monday, August 28, 2017 11:16 AM
> > To: Eduardo Otubo <[email protected]>
> > Cc: [email protected]; [email protected]; Haiyang
> > Zhang <[email protected]>; Stephen Hemminger
> > <[email protected]>; David Miller <[email protected]>
> > Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
> >
> > On Mon, 28 Aug 2017 12:01:21 +0200
> > Eduardo Otubo <[email protected]> wrote:
> >
> > > v2: The script is now a little bit safer so it doesn't conflicts with
> > > network daemon trying to set configurations at the same time.
> > >
> > > This patch fixes the behavior of the hv_set_ifconfig script when
> > setting
> > > the interface ip. Sometimes the interface has already been configured
> > by
> > > network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> > > exists error"; in order to avoid this error this patch makes sure
> > double
> > > checks the interface before trying anything.
> > >
> > > Signed-off-by: Eduardo Otubo <[email protected]>
> >
> > Adding new dependency on systemd is not going to make this script
> > even less useful. I wonder why the script still exists at all? Most of
> > the
> > Linux distro's can already setup HV networking without it.
> >
Also, this script is meant to run only on RHEL guests, as written on
the header of this very same file.
--
Eduardo Otubo
Senior Software Engineer @ RedHat
> -----Original Message-----
> From: Haiyang Zhang
> Sent: Monday, August 28, 2017 8:57 AM
> To: Stephen Hemminger <[email protected]>; Eduardo Otubo
> <[email protected]>; KY Srinivasan <[email protected]>
> Cc: [email protected]; [email protected]; Stephen
> Hemminger <[email protected]>; David Miller
> <[email protected]>
> Subject: RE: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
>
>
>
> > -----Original Message-----
> > From: Stephen Hemminger [mailto:[email protected]]
> > Sent: Monday, August 28, 2017 11:16 AM
> > To: Eduardo Otubo <[email protected]>
> > Cc: [email protected]; [email protected]; Haiyang
> > Zhang <[email protected]>; Stephen Hemminger
> > <[email protected]>; David Miller <[email protected]>
> > Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
> >
> > On Mon, 28 Aug 2017 12:01:21 +0200
> > Eduardo Otubo <[email protected]> wrote:
> >
> > > v2: The script is now a little bit safer so it doesn't conflicts with
> > > network daemon trying to set configurations at the same time.
> > >
> > > This patch fixes the behavior of the hv_set_ifconfig script when
> > setting
> > > the interface ip. Sometimes the interface has already been configured
> > by
> > > network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> > > exists error"; in order to avoid this error this patch makes sure
> > double
> > > checks the interface before trying anything.
> > >
> > > Signed-off-by: Eduardo Otubo <[email protected]>
> >
> > Adding new dependency on systemd is not going to make this script
> > even less useful. I wonder why the script still exists at all? Most of
> > the
> > Linux distro's can already setup HV networking without it.
> >
>
> This script is used by a host to inject IP into guests. KY knows more
> details about it.
I wrote this script initially to provide an example script for Distros to base their solution on.
KVP supports IP injection to enable VM migration. For this scenario, I think we recommend that NM be
disabled.
K. Y
>
On Mon, Aug 28, 2017 at 04:48:32PM +0000, KY Srinivasan wrote:
>
>
> > -----Original Message-----
> > From: Haiyang Zhang
> > Sent: Monday, August 28, 2017 8:57 AM
> > To: Stephen Hemminger <[email protected]>; Eduardo Otubo
> > <[email protected]>; KY Srinivasan <[email protected]>
> > Cc: [email protected]; [email protected]; Stephen
> > Hemminger <[email protected]>; David Miller
> > <[email protected]>
> > Subject: RE: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
> >
> >
> >
> > > -----Original Message-----
> > > From: Stephen Hemminger [mailto:[email protected]]
> > > Sent: Monday, August 28, 2017 11:16 AM
> > > To: Eduardo Otubo <[email protected]>
> > > Cc: [email protected]; [email protected]; Haiyang
> > > Zhang <[email protected]>; Stephen Hemminger
> > > <[email protected]>; David Miller <[email protected]>
> > > Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
> > >
> > > On Mon, 28 Aug 2017 12:01:21 +0200
> > > Eduardo Otubo <[email protected]> wrote:
> > >
> > > > v2: The script is now a little bit safer so it doesn't conflicts with
> > > > network daemon trying to set configurations at the same time.
> > > >
> > > > This patch fixes the behavior of the hv_set_ifconfig script when
> > > setting
> > > > the interface ip. Sometimes the interface has already been configured
> > > by
> > > > network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> > > > exists error"; in order to avoid this error this patch makes sure
> > > double
> > > > checks the interface before trying anything.
> > > >
> > > > Signed-off-by: Eduardo Otubo <[email protected]>
> > >
> > > Adding new dependency on systemd is not going to make this script
> > > even less useful. I wonder why the script still exists at all? Most of
> > > the
> > > Linux distro's can already setup HV networking without it.
> > >
> >
> > This script is used by a host to inject IP into guests. KY knows more
> > details about it.
>
> I wrote this script initially to provide an example script for Distros to base their solution on.
> KVP supports IP injection to enable VM migration. For this scenario, I think we recommend that NM be
> disabled.
>
So, what you're saying is that this should be fixed downstream,
instead? This solution seems pretty safe for me and long term we can
think about something else that could get rid of this script. So NM or
whatever is in use can actually do the configuration.
Any chance to have this patch ACK'd as a form of a short term
solution?
> -----Original Message-----
> From: Eduardo Otubo [mailto:[email protected]]
> Sent: Thursday, August 31, 2017 1:17 AM
> To: KY Srinivasan <[email protected]>
> Cc: Haiyang Zhang <[email protected]>; Stephen Hemminger
> <[email protected]>; [email protected];
> [email protected]; Stephen Hemminger
> <[email protected]>; David Miller <[email protected]>
> Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
>
> On Mon, Aug 28, 2017 at 04:48:32PM +0000, KY Srinivasan wrote:
> >
> >
> > > -----Original Message-----
> > > From: Haiyang Zhang
> > > Sent: Monday, August 28, 2017 8:57 AM
> > > To: Stephen Hemminger <[email protected]>; Eduardo
> Otubo
> > > <[email protected]>; KY Srinivasan <[email protected]>
> > > Cc: [email protected]; [email protected]; Stephen
> > > Hemminger <[email protected]>; David Miller
> > > <[email protected]>
> > > Subject: RE: [PATCHv2] hv_set_ifconfig.sh double check before setting ip
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: Stephen Hemminger [mailto:[email protected]]
> > > > Sent: Monday, August 28, 2017 11:16 AM
> > > > To: Eduardo Otubo <[email protected]>
> > > > Cc: [email protected]; [email protected];
> Haiyang
> > > > Zhang <[email protected]>; Stephen Hemminger
> > > > <[email protected]>; David Miller <[email protected]>
> > > > Subject: Re: [PATCHv2] hv_set_ifconfig.sh double check before setting
> ip
> > > >
> > > > On Mon, 28 Aug 2017 12:01:21 +0200
> > > > Eduardo Otubo <[email protected]> wrote:
> > > >
> > > > > v2: The script is now a little bit safer so it doesn't conflicts with
> > > > > network daemon trying to set configurations at the same time.
> > > > >
> > > > > This patch fixes the behavior of the hv_set_ifconfig script when
> > > > setting
> > > > > the interface ip. Sometimes the interface has already been
> configured
> > > > by
> > > > > network daemon, in this case hv_set_ifconfig causes "RTNETLINK: file
> > > > > exists error"; in order to avoid this error this patch makes sure
> > > > double
> > > > > checks the interface before trying anything.
> > > > >
> > > > > Signed-off-by: Eduardo Otubo <[email protected]>
> > > >
> > > > Adding new dependency on systemd is not going to make this script
> > > > even less useful. I wonder why the script still exists at all? Most of
> > > > the
> > > > Linux distro's can already setup HV networking without it.
> > > >
> > >
> > > This script is used by a host to inject IP into guests. KY knows more
> > > details about it.
> >
> > I wrote this script initially to provide an example script for Distros to base
> their solution on.
> > KVP supports IP injection to enable VM migration. For this scenario, I think
> we recommend that NM be
> > disabled.
> >
>
> So, what you're saying is that this should be fixed downstream,
> instead? This solution seems pretty safe for me and long term we can
> think about something else that could get rid of this script. So NM or
> whatever is in use can actually do the configuration.
>
> Any chance to have this patch ACK'd as a form of a short term
> solution?
Given that this is just an example script to be customized on a per Distro basis, do
you still see value in making this change upstream. In any case, as Olaf noted the delay will
cause problems. One option would be to simply fail if network scripts are running. As you say, if you
are making sure that the KVP daemon is only run after the network script have executed, this restriction
(failing if the scripts are active) should not be an issue.
Regards,
K. Y