- Devel - libvirt List Archives

[libvirt] [PATCH] Fix trailing whitespace in NEWS
by Mark McLoughlin 29 Jul '09

29 Jul '09

Hey, I had to commit this before I could push Cheers, Mark.

1 0

[libvirt] [PATCH 00/14] Clean up spec files mess
by Mark McLoughlin 29 Jul '09

29 Jul '09

Hey, I get treated like a war criminal when I make a change in the Fedora spec file and don't send it upstream, yet there are a whole heap of changes which haven't been pushed upstream ... only one of which is mine ... two changes less than DV himself :-) The following series of patches pulls in the changes from Fedora which make sense. Some of them I'm not 100% sure of: - defaulting to policy kit - is there some reason we don't want that upstream? - The python cruft thing, not sure its needed - Converting NEWS to UTF-8 could probably be done upstream - Requiring libselinux is dubious, doesn't autorequires handle that? My final contention is that maintaining these spec files in the upstream repo actually makes it harder to maintain the Fedora spec file and helps nobody. Because of half-assed attempts to keep the two in sync, I ended up having to fix a bunch of stuff and re-instate another bunch of stuff in Fedora: - Enable netcf support - Pass --with-qemu-user=qemu etc. to configure - Move various requires to the libvirt-client sub-package - Sync some trivial cleanups from upstream spec file - Remove explicit libxml2 requires, again - Build with --without-capng if capng support is disabled - Remove explicit dir creating in makeinstall, replaced by attr in files - Set perms on /var/{run,lib,cache}/libvirt/qemu I really think we should kill these off and just point people at the Fedora spec file if they're looking for a reference spec file. I need some very strong rationale before I ever go wasting my time (2 hours) on this again. Cheers, Mark.

4 23

[libvirt] [PATCH] Remove MAX_TAP_ID
by Aron Griffis 29 Jul '09

29 Jul '09

As far as I can tell, there's no reason to format the device string in brAddTap(). Delegate the job to TUNSETIFF, thereby removing the loop and the MAX_TAP_ID artificial limit. This patch allows me to get 421 guests running before hitting other limits. Signed-off-by: Aron Griffis <aron.griffis(a)hp.com> --- src/bridge.c | 94 +++++++++++++++++++++------------------------------------- 1 files changed, 34 insertions(+), 60 deletions(-) diff --git a/.gnulib b/.gnulib deleted file mode 160000 index b653eda..0000000 --- a/.gnulib +++ /dev/null @@ -1 +0,0 @@ -Subproject commit b653eda3ac4864de205419d9f41eec267cb89eeb diff --git a/src/bridge.c b/src/bridge.c index 0509afd..ec37192 100644 --- a/src/bridge.c +++ b/src/bridge.c @@ -49,8 +49,6 @@ #include "util.h" #include "logging.h" -#define MAX_TAP_ID 256 - #define JIFFIES_TO_MS(j) (((j)*1000)/HZ) #define MS_TO_JIFFIES(ms) (((ms)*HZ)/1000) @@ -466,76 +464,52 @@ brAddTap(brControl *ctl, int vnet_hdr, int *tapfd) { - int id, subst, fd; + int fd, len; + struct ifreq ifr = {0}; if (!ctl || !ctl->fd || !bridge || !ifname) return EINVAL; - subst = id = 0; - - if (strstr(*ifname, "%d")) - subst = 1; - if ((fd = open("/dev/net/tun", O_RDWR)) < 0) return errno; - if (vnet_hdr) - vnet_hdr = brProbeVnetHdr(fd); + ifr.ifr_flags = IFF_TAP|IFF_NO_PI; - do { - struct ifreq try; - int len; +#ifdef IFF_VNET_HDR + if (vnet_hdr && brProbeVnetHdr(fd)) + ifr.ifr_flags |= IFF_VNET_HDR; +#endif - memset(&try, 0, sizeof(struct ifreq)); + strncpy(ifr.ifr_name, *ifname, IFNAMSIZ-1); - try.ifr_flags = IFF_TAP|IFF_NO_PI; + if (ioctl(fd, TUNSETIFF, &ifr) < 0) + goto error; -#ifdef IFF_VNET_HDR - if (vnet_hdr) - try.ifr_flags |= IFF_VNET_HDR; -#endif + len = strlen(ifr.ifr_name); + if (len >= BR_IFNAME_MAXLEN - 1) { + errno = EINVAL; + goto error; + } - if (subst) { - len = snprintf(try.ifr_name, BR_IFNAME_MAXLEN, *ifname, id); - if (len >= BR_IFNAME_MAXLEN) { - errno = EADDRINUSE; - goto error; - } - } else { - len = strlen(*ifname); - if (len >= BR_IFNAME_MAXLEN - 1) { - errno = EINVAL; - goto error; - } - - strncpy(try.ifr_name, *ifname, len); - try.ifr_name[len] = '\0'; - } - - if (ioctl(fd, TUNSETIFF, &try) == 0) { - /* We need to set the interface MTU before adding it - * to the bridge, because the bridge will have its - * MTU adjusted automatically when we add the new interface. - */ - if ((errno = brSetInterfaceMtu(ctl, bridge, try.ifr_name))) - goto error; - if ((errno = brAddInterface(ctl, bridge, try.ifr_name))) - goto error; - if ((errno = brSetInterfaceUp(ctl, try.ifr_name, 1))) - goto error; - if (!tapfd && - (errno = ioctl(fd, TUNSETPERSIST, 1))) - goto error; - VIR_FREE(*ifname); - if (!(*ifname = strdup(try.ifr_name))) - goto error; - if (tapfd) - *tapfd = fd; - return 0; - } - - id++; - } while (subst && id <= MAX_TAP_ID); + /* We need to set the interface MTU before adding it + * to the bridge, because the bridge will have its + * MTU adjusted automatically when we add the new interface. + */ + if ((errno = brSetInterfaceMtu(ctl, bridge, ifr.ifr_name))) + goto error; + if ((errno = brAddInterface(ctl, bridge, ifr.ifr_name))) + goto error; + if ((errno = brSetInterfaceUp(ctl, ifr.ifr_name, 1))) + goto error; + if (!tapfd && + (errno = ioctl(fd, TUNSETPERSIST, 1))) + goto error; + VIR_FREE(*ifname); + if (!(*ifname = strdup(ifr.ifr_name))) + goto error; + if (tapfd) + *tapfd = fd; + return 0; error: close(fd);

1 0

[libvirt] ANNOUNCE: New release virtinst 0.500.0
by Cole Robinson 29 Jul '09

29 Jul '09

I'm happy to announce a new virtinst release, version 0.500.0. The release can be downloaded from: http://virt-manager.org/download.html The direct download link is: http://virt-manager.org/download/sources/virtinst/virtinst-0.500.0.tar.gz This release includes: - New virt-install device options --serial, --parallel, and --video - Allow various auth types for libvirt connections (PolicyKit, SASL, ...) - New virt-clone option --auto-clone: generates all needed input. - Option to specify network device model via virt-install --network (Guido Gunther) - New virt-install option --virt-type for specifying hypervisor type (kvm, qemu). --accelerate is now the default behavior: To provision a plain qemu VM on a KVM enabled host, use '--virt-type qemu' - OVF input support for virt-convert - Many bug fixes and improvements Thanks to everyone who has contributed to this release through testing, bug reporting, submitting patches, and otherwise sending in feedback! Thanks, Cole

1 0

[libvirt] [PATCH] qemu: fix missing ifname in qemu argument
by Ryota Ozaki 28 Jul '09

28 Jul '09

Hi, This patch adds missing ifname, which is a tap name for network/bridge configuration, in qemu argument of -net. Without this patch, a specified tap name, e.g., <target dev='tap01'/>, in a domain XML will be ignored. Thanks, ozaki-r >From feeca4f2cecf5c6836f7aec6e004585f3b855bed Mon Sep 17 00:00:00 2001 From: Ryota Ozaki <ozaki.ryota(a)gmail.com> Date: Tue, 28 Jul 2009 23:46:12 +0900 Subject: [PATCH] qemu: fix missing ifname in qemu argument --- src/qemu_conf.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/qemu_conf.c b/src/qemu_conf.c index 4fe4e39..48bf928 100644 --- a/src/qemu_conf.c +++ b/src/qemu_conf.c @@ -1187,11 +1187,13 @@ qemuBuildHostNetStr(virConnectPtr conn, switch (net->type) { case VIR_DOMAIN_NET_TYPE_NETWORK: case VIR_DOMAIN_NET_TYPE_BRIDGE: - if (virAsprintf(str, "%stap%cfd=%s,vlan=%d%s%s", + if (virAsprintf(str, "%stap%cfd=%s,vlan=%d%s%s%s%s", prefix ? prefix : "", type_sep, tapfd, vlan, (net->hostnet_name ? ",name=" : ""), - (net->hostnet_name ? net->hostnet_name : "")) < 0) { + (net->hostnet_name ? net->hostnet_name : ""), + (net->ifname ? ",ifname=" : ""), + (net->ifname ? net->ifname : "")) < 0) { virReportOOMError(conn); return -1; } -- 1.6.0.6

2 2

[libvirt] Java JNA Patch series
by Bryan Kearney 28 Jul '09

28 Jul '09

Per the earlier email, here is the patch series

2 32

Re: [libvirt] VMware support and libcurl on rhel-u1
by Shahar Klein 28 Jul '09

28 Jul '09

apperantly I needed to unset CURLOPT_SSL_VERIFYHOST as well patched that and it is working now 10x Shahar ________________________________ From: Tom Hughes <tom(a)compton.nu> To: Shahar Klein <shaharklein(a)yahoo.com> Sent: Monday, July 27, 2009 2:36:32 PM Subject: Re: [libvirt] VMware support and libcurl on rhel-u1 On 27/07/09 12:33, Shahar Klein wrote: > this is what I thought but it doesn't work for me: > > [root@rain8 libvirt]# virsh -c esx://rain3?transport=http > Enter username for rain3 [root]: > Enter root password for rain3: > error: internal error curl_easy_perform() returned an error: SSL peer > certificate was not ok (51) > error: failed to connect to the hypervisor > > maybe because I'm using curl 7.15? That looks like it is still checking the certificate. Are you sure you're not still running the unpatched one somehow? I'm running curl 7.19.4 which isn't that different - does the curl_easy_setopt manual page mention the CURLOPT_SSL_VERIFYPEER option? Tom -- Tom Hughes (tom(a)compton.nu) http://www.compton.nu/

4 5

[libvirt] [PATCH 2/2] RPM spec file updated with glusterfs dependency
by Harshavardhana 28 Jul '09

28 Jul '09

Add new dependency for glusterfs rpm. --- libvirt.spec.in | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/libvirt.spec.in b/libvirt.spec.in index 1e8a8ef..ccef3a5 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -86,6 +86,8 @@ BuildRequires: util-linux # For showmount in FS driver (netfs discovery) BuildRequires: nfs-utils Requires: nfs-utils +# For glusterfs +Requires: glusterfs-client >= 2.0.2 %endif %if %{with_qemu} # From QEMU RPMs -- 1.6.0.6

4 6

[libvirt] [RFC/Experimental]: Tunnelled migration
by Chris Lalancette 28 Jul '09

28 Jul '09

All, Attached is the current version of the tunnelled migration patch, based upon danpb's generic datastream work. In order to use this work, you must first grab danpb's data-streams git branch here: http://gitorious.org/~berrange/libvirt/staging and then apply this patch on top. In some basic testing, this seems to work fine for me, although I have not given it a difficult scenario nor measured CPU utilization with these patches in place. DanB, these patches take a slightly different approach than you and I discussed yesterday on IRC. Just to recap, you suggested a new version of virMigratePrepare (called virMigratePrepareTunnel) that would take in as one of the arguments a datastream, and during the prepare step properly setup the datastream. Unless I'm missing something (which is entirely possible), this would also require passing that same datastream into the perform and finish stages, meaning that I'd essentially have an all new migration protocol version 3. To try to avoid that, during the prepare I store the port that we used to start the listening qemu in a new field in the virDomainObj structure. Then during the perform step, I create a datastream on the destination and run a new RPC function called virDomainMigratePrepareTunnel. This looks that port back up, associates it with the current stream, and returns back to the caller. Then the source side just does virStreamSend for all the data, and we have tunnelled migration. TODO: - More testing, especially under worst-case scenarios (VM constantly changing it's memory during migration) - CPU utilization testing to make sure that we aren't using a lot of CPU time doing this - Wall-clock testing - Switch over to using Unix Domain Sockets instead of localhost TCP migration. With a patch I put into upstream qemu (and is now in F-12), we can totally get rid of the scanning of localhost ports to find a free one, and just use Unix Domain Sockets. That should make the whole thing more robust. -- Chris Lalancette

2 2

[libvirt] test driver and virtOpenAuth()
by Bryan Kearney 28 Jul '09

28 Jul '09

I am trying to test a call into virConnectOpenAuth. How can I configure the test driver to accept auth calls on "test+tcp://localhost/default" Thanks! -- bk

2 4