- Devel - libvirt List Archives

[libvirt] [PATCH 00/15] Rewrite the QEMU monitor handling v2
by Daniel P. Berrange 05 Nov '09

05 Nov '09

This is an update of this series: http://www.redhat.com/archives/libvir-list/2009-October/msg00644.html minus the patches DV already ACKd which I've merged. The main change in this series is that I've removed the change which added a RW-lock primitive in the QEMU driver. Instead I've implemented the job condition variable idea I described here http://www.redhat.com/archives/libvir-list/2009-October/msg00815.html The nice thing about this is that its actually possible to start doing timeouts of commands. This series only partially supports timeouts. A long running monitor command itself will not timeout, but if a second API call is made while someone else is using the monitor, this waiting call will timeout after 30 seconds if the monitor was not relased. Daniel

3 37

[libvirt] [PATCH] network utilities: Add functions for address->text and get/set port number
by Matthew Booth 05 Nov '09

05 Nov '09

--- src/libvirt_private.syms | 3 ++ src/util/network.c | 88 +++++++++++++++++++++++++++++++++++++++++++++- src/util/network.h | 6 +++ 3 files changed, 96 insertions(+), 1 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 8525dbd..15d75fd 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -289,10 +289,13 @@ virFree; virSocketAddrInNetwork; virSocketAddrIsNetmask; virSocketCheckNetmask; +virSocketFormatAddr; +virSocketGetPort; virSocketGetRange; virSocketParseAddr; virSocketParseIpv4Addr; virSocketParseIpv6Addr; +virSocketSetPort; # network_conf.h diff --git a/src/util/network.c b/src/util/network.c index abd866c..094130f 100644 --- a/src/util/network.c +++ b/src/util/network.c @@ -9,6 +9,7 @@ */ #include <config.h> +#include <arpa/inet.h> #include "memory.h" #include "network.h" @@ -64,7 +65,7 @@ static int getIPv6Addr(virSocketAddrPtr addr, virIPv6AddrPtr tab) { * Mostly a wrapper for getaddrinfo() extracting the address storage * from the numeric string like 1.2.3.4 or 2001:db8:85a3:0:0:8a2e:370:7334 * - * Returns the lenght of the network address or -1 in case of error. + * Returns the length of the network address or -1 in case of error. */ int virSocketParseAddr(const char *val, virSocketAddrPtr addr, int hint) { @@ -116,6 +117,91 @@ virSocketParseIpv6Addr(const char *val, virSocketAddrPtr addr) { return(virSocketParseAddr(val, addr, AF_INET6)); } +/* + * virSocketFormatAddr: + * @addr: an initialised virSocketAddrPtr + * + * Returns a string representation of the given address + * Returns NULL on any error + * Caller must free the returned string + */ +char * +virSocketFormatAddr(virSocketAddrPtr addr) { + char *out; + size_t outlen; + void *inaddr; + + if (addr->stor.ss_family == AF_INET) { + outlen = INET_ADDRSTRLEN; + inaddr = &addr->inet4.sin_addr; + } + + else if (addr->stor.ss_family == AF_INET6) { + outlen = INET6_ADDRSTRLEN; + inaddr = &addr->inet6.sin6_addr; + } + + else { + return NULL; + } + + if (VIR_ALLOC_N(out, outlen) < 0) + return NULL; + + if (inet_ntop(addr->stor.ss_family, inaddr, out, outlen) == NULL) { + VIR_FREE(out); + return NULL; + } + + return out; +} + +/* + * virSocketSetPort: + * @addr: an initialised virSocketAddrPtr + * @port: the port number to set + * + * Set the transport layer port of the given virtSocketAddr + * + * Returns 0 on success, -1 on failure + */ +int +virSocketSetPort(virSocketAddrPtr addr, in_port_t port) { + if(addr->stor.ss_family == AF_INET) { + addr->inet4.sin_port = port; + } + + else if(addr->stor.ss_family == AF_INET6) { + addr->inet6.sin6_port = port; + } + + else { + return -1; + } + + return 0; +} + +/* + * virSocketGetPort: + * @addr: an initialised virSocketAddrPtr + * + * Returns the transport layer port of the given virtSocketAddr + * Returns 0 if @addr is invalid + */ +in_port_t +virSocketGetPort(virSocketAddrPtr addr) { + if(addr->stor.ss_family == AF_INET) { + return addr->inet4.sin_port; + } + + else if(addr->stor.ss_family == AF_INET6) { + return addr->inet6.sin6_port; + } + + return 0; +} + /** * virSocketAddrIsNetmask: * @netmask: the netmask address diff --git a/src/util/network.h b/src/util/network.h index e590747..7618547 100644 --- a/src/util/network.h +++ b/src/util/network.h @@ -34,6 +34,12 @@ int virSocketParseIpv4Addr(const char *val, int virSocketParseIpv6Addr(const char *val, virSocketAddrPtr addr); +char * virSocketFormatAddr(virSocketAddrPtr addr); + +int virSocketSetPort(virSocketAddrPtr addr, in_port_t port); + +in_port_t virSocketGetPort(virSocketAddrPtr addr); + int virSocketAddrInNetwork(virSocketAddrPtr addr1, virSocketAddrPtr addr2, virSocketAddrPtr netmask); -- 1.6.2.5

3 7

[libvirt] [PATCH] create() and destroy() support for Power Hypervisor
by Eduardo Otubo 05 Nov '09

05 Nov '09

Hello all, This patch includes a lot of changes: * I changed all references of uuid_db to uuid_table. Bigger name, but this semantic has a better understanding. * Now we have a little control of UUID generated for each partition. It's based on a table that matches UUID with LPAR's IDs, I keep it in a file that is stored in the managed system (HMC or IVM). Even having isolated functions to manipulate this file I still need to implement a way to lock it and make the operation atomic to avoid corruptions. * The XML file used in the create() function still must be improved. I used the <disk> tag to make a work around to handle storage pools. Now I ask for some help, how do I use the <pool> tag if there is no reference to it at the virDomainDef structure? I've been told that libvirt possibly would make a mini-release this week to push some major fixes on Fedora 12. Is there a little possibility to push these new phyp features too? This would be very important for the phyp project, since this functions are not just to manage partitions but manipulate them. Any comments are always welcome. []'s -- Eduardo Otubo Software Engineer Linux Technology Center IBM Systems & Technology Group Mobile: +55 19 8135 0885 eotubo(a)linux.vnet.ibm.com

5 13

[libvirt] Remote access and libvirtd
by Dave Bryson 05 Nov '09

05 Nov '09

I need to be able to access libvirt remotely. Reading through the documention is says "libvirtd should be running on the remote machine" for clarification does this mean that libvirtd needs to run *on* the remote hypervisor? Thanks, Dave

2 1

[libvirt] Add support for legacy -vmchannel option
by Matthew Booth 05 Nov '09

05 Nov '09

These patches add support for the legacy qemu -vmchannel option. The following domain xml: <channel type='unix'> <source mode='bind' path='/tmp/vmchannel'/> <target type='vmchannel' deviceid='0200'/> </channel> becomes: -vmchannel di:0200,unix:/tmp/vmchannel,server,nowait I don't expect these to be merged, but I would appreciate a review. Thanks, Matt

4 6

[libvirt] Guestfwd support (round 3)
by Matthew Booth 05 Nov '09

05 Nov '09

This is substantially the same as the last set. I don't think I changed patches 1 and 2 at all. Patch 3 is updated to use a virBuffer as Rich Jones recommended. Patch 4 is updated slightly for the above change, and there are a couple of other trivial tidy ups in there. Patch 5 is new. QEMU's guestfwd only supports IPv4 addresses, so we check for that. Rich Jones also noted that guestfwd is specific to QEMU. If there's a good chance this same syntax might do something simililar in a different driver, I agree that changing the name would be a good idea. However, I'm not familiar enough with other hypervisors to be able to make a call on this. If nothing else does anything similar, I'd keep the QEMU name for clarity. Matt

3 17

[libvirt] [PATCH] LXC allow container to have ethN interfaces
by Ryota Ozaki 05 Nov '09

05 Nov '09

Current implementation of lxc driver creates vethN named interface(s) in the host and passes as it is to a container. The reason why it doesn't use ethN is due to the limitation that one namespace cannot have multiple iterfaces that have an identical name so that we give up creating ethN named interface in the host for the container. However, we should be able to allow the container to have ethN by changing the name after clone(CLONE_NEWNET). The patch simply does that and numberes to ethN with ascending order from zero with no concern for what vethN is named. So say if there is one interface, its name will be eth0. Note that this patch is based on top of the prior cleanup patches. --- src/lxc/lxc_container.c | 34 ++++++++++++++++++++++++---------- src/lxc/veth.c | 31 +++++++++++++++++++++++++++++++ src/lxc/veth.h | 1 + 3 files changed, 56 insertions(+), 10 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 2419345..97b7903 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -225,26 +225,38 @@ static int lxcContainerWaitForContinue(int control) /** - * lxcContainerEnableInterfaces: + * lxcContainerRenameAndEnableInterfaces: * @nveths: number of interfaces * @veths: interface names * - * This function will enable the interfaces for this container. + * This function will rename the interfaces to ethN + * with id ascending order from zero and enable the + * renamed interfaces for this container. * * Returns 0 on success or nonzero in case of error */ -static int lxcContainerEnableInterfaces(unsigned int nveths, - char **veths) +static int lxcContainerRenameAndEnableInterfaces(unsigned int nveths, + char **veths) { int rc = 0; unsigned int i; + char *newname = NULL; for (i = 0 ; i < nveths ; i++) { - DEBUG("Enabling %s", veths[i]); - rc = vethInterfaceUpOrDown(veths[i], 1); - if (0 != rc) { + rc = virAsprintf(&newname, "eth%d", i); + if (rc < 0) goto error_out; - } + + DEBUG("Renaming %s to %s", veths[i], newname); + rc = setInterfaceName(veths[i], newname); + if (0 != rc) + goto error_out; + + DEBUG("Enabling %s", newname); + rc = vethInterfaceUpOrDown(newname, 1); + if (0 != rc) + goto error_out; + VIR_FREE(newname); } /* enable lo device only if there were other net devices */ @@ -252,6 +264,7 @@ static int lxcContainerEnableInterfaces(unsigned int nveths, rc = vethInterfaceUpOrDown("lo", 1); error_out: + VIR_FREE(newname); return rc; } @@ -757,8 +770,9 @@ static int lxcContainerChild( void *data ) if (lxcContainerWaitForContinue(argv->monitor) < 0) return -1; - /* enable interfaces */ - if (lxcContainerEnableInterfaces(argv->nveths, argv->veths) < 0) + /* rename and enable interfaces */ + if (lxcContainerRenameAndEnableInterfaces(argv->nveths, + argv->veths) < 0) return -1; /* drop a set of root capabilities */ diff --git a/src/lxc/veth.c b/src/lxc/veth.c index 8617cf7..ede85ce 100644 --- a/src/lxc/veth.c +++ b/src/lxc/veth.c @@ -247,3 +247,34 @@ int setMacAddr(const char* iface, const char* macaddr) error_out: return rc; } + +/** + * setInterfaceName + * @iface: name of device + * @new: new name of @iface + * + * Changes the name of the given device with the + * given new name using this command: + * ip link set @iface name @new + * + * Returns 0 on success or -1 in case of error + */ +int setInterfaceName(const char* iface, const char* new) +{ + int rc = -1; + const char *argv[] = { + "ip", "link", "set", iface, "name", new, NULL + }; + int cmdResult; + + if (NULL == iface || NULL == new) { + goto error_out; + } + + rc = virRun(NULL, argv, &cmdResult); + if (0 == rc) + rc = cmdResult; + +error_out: + return rc; +} diff --git a/src/lxc/veth.h b/src/lxc/veth.h index 8f2f514..8075a5e 100644 --- a/src/lxc/veth.h +++ b/src/lxc/veth.h @@ -21,5 +21,6 @@ int vethDelete(const char* veth); int vethInterfaceUpOrDown(const char* veth, int upOrDown); int moveInterfaceToNetNs(const char *iface, int pidInNs); int setMacAddr(const char* iface, const char* macaddr); +int setInterfaceName(const char* iface, const char* new); #endif /* VETH_H */ -- 1.6.2.5

2 1

[libvirt] [PATCH 0/3] LXC misc fixes and cleanups
by Ryota Ozaki 05 Nov '09

05 Nov '09

This patch series fix and cleanup over whole lxc codes. I want to clean up before adding more in them. Thanks, ozaki-r Ryota Ozaki (3): LXC fix wrong or out-of-date function descriptions LXC cleanup and fix lxcError LXC cleanup deep indentation in lxcDomainSetAutostart src/lxc/lxc_container.c | 85 +++++++++++----------- src/lxc/lxc_controller.c | 40 ++++++----- src/lxc/lxc_driver.c | 177 +++++++++++++++++++++++++--------------------- src/lxc/veth.c | 8 +- 4 files changed, 164 insertions(+), 146 deletions(-)

3 7

[libvirt] flush guest page cache and suspend?
by Marco Colombo 05 Nov '09

05 Nov '09

Hi, I'm new to this list. I've searched the archives (and google) but I haven't found an answer. Recently I moved to CentOS 5.4, with libvirt-0.6.3 and kvm-83. They are rather new compared to 5.3 ones, so I'm still investigating on the new capabilities. My guests disk images are LV single devices, /dev/vg_virt/<guestname>, usually with two partitions, the root fs and swap space. I'd like to run backups on the node only, taking snapshots of the LVs. I'm already doing that, actually. Everything works fine, using bacula (with one minor glitch, but that's for another list). Now I wonder if it makes sense to suspend the guest before creating the snapshot: virsh suspend <guestname> lvcreate --snapshot ... virsh resume <guestname> ... perform backup ... lvremove -f <the snapshot> but thinking about it a second time it seems it doesn't add anything to the equation, the filesystem is still potentially unclean, with data to be written still in the guest page cache. The filesystem is ext3, so it's not that bad (recovers a consistent state from the journal, with standard options even data-consistent, not just metadata-consistent). Yet it's like doing a backup of a (potentially badly) crashed filesystem. While I trust the recovery step of ext3, my feeling is that it can't be 100% reliable. I have (rarely) seen ext3 crash and ask for fsck at boot, journal or not (for truth's sake, I can't rule out an hard disk problem in those cases). So, first question: does the suspend command cause a flush in the guest OS (details: both guest and node are CentOS 5.4, hypervisor is qemu-kvm)? I guess not (otherwise I won't be here). So if not, what are the options to force the sync? Ideally, there should be a single atomic 'sync & suspend'. In practice, I can think of some workarounds: ssh <guest> -c sync, or the very old-fashioned way of enabling the 'sync' user and logging in from the serial console, or issuing a sysrq-s, again on the console. I'm interested in the latter, but I wasn't able to trigger the sysrq from either 'virsh console <guestname>' or 'screen `virsh ttyconsole <guestname>` (tried minicom, too). The serial console is on pty, I'm not even sure you can generate a break on a pty. (and yes, I remembered to sysctl -w kernel.sysrq=1 on the guest). I know XEN has 'xm sysrq', but this is kvm. Is there anything similar? I think it can be done if you invoke qemu-kvm from the command line with -nographics (it multiplexes the console and the monitor lines on stdin/out, and you can send a 'break' with C-a b I think, I've never tried). So question 2): is there a way to send a sysrq-s to the guest? My fallback plan is to try and map the serial line over telnet instead of pty, and the figure out a way to send the break (which I think is part of the telnet protocol) from the command line. I'd rather not mess with telnet and local port assignment to the guests, if not necessary. My really fallback plan is to revert to shutdown/snapshot/create, which is overkill for a daily backup. Question 3): does it _really_ make sense to try and sync the guest OS page cache before taking the snapshot? Or is it just me being paranoid? For reference, here's the qemu cmdline of one of the guests: /usr/libexec/qemu-kvm -S -M pc -m 512 -smp 1 -name wtitv -uuid 445d0fe5-c1b6-4baf-a186-da1fe021158c -monitor pty -pidfile /var/run/libvirt/qemu//wtitv.pid -boot c -drive file=/dev/vg_virt/wtitv,if=ide,index=0,boot=on -net nic,macaddr=00:16:3e:57:45:4f,vlan=0,model=e1000 -net tap,fd=14,script=,vlan=0,ifname=vnet0 -serial pty -parallel none -usb -usbdevice tablet -vnc 127.0.0.1:0 and here's the xml config file for libvirt, same guest: <domain type='kvm'> <name>wtitv</name> <uuid>445d0fe5-c1b6-4baf-a186-da1fe021158c</uuid> <memory>524288</memory> <currentMemory>524288</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64' machine='pc'>hvm</type> <boot dev='hd'/> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <features> <acpi/> </features> <devices> <emulator>/usr/libexec/qemu-kvm</emulator> <disk type='block' device='disk'> <source dev='/dev/vg_virt/wtitv'/> <target dev='hda' bus='ide'/> </disk> <interface type='bridge'> <mac address='00:16:3e:57:45:4f'/> <source bridge='br1'/> <model type='e1000'/> </interface> <console type='pty'> <target port='0'/> </console> <input type='tablet' bus='usb'/> <graphics type='vnc' autoport='yes'/> </devices> </domain> TIA, .TM.

2 2

[libvirt] [PATCH 0/2] CPU flags capabilities and domain configuration
by Jiri Denemark 05 Nov '09

05 Nov '09

Hi, These are two initial patches for CPU flags/masks support for libvirt. The first patch contains changes to schemas and describes semantics of new elements and attributes. The second patch introduces internal API functions for XML parsing and formating. Jiri Denemark (2): XML schema for CPU flags XML parsing/formating code for CPU flags docs/schemas/capability.rng | 30 ++++ docs/schemas/domain.rng | 49 ++++++ include/libvirt/virterror.h | 1 + src/Makefile.am | 9 +- src/conf/capabilities.c | 31 ++++- src/conf/capabilities.h | 6 + src/conf/cpu_conf.c | 345 +++++++++++++++++++++++++++++++++++++++++++ src/conf/cpu_conf.h | 110 ++++++++++++++ src/conf/domain_conf.c | 15 ++ src/conf/domain_conf.h | 2 + src/libvirt_private.syms | 8 + src/util/virterror.c | 3 + 12 files changed, 606 insertions(+), 3 deletions(-) create mode 100644 src/conf/cpu_conf.c create mode 100644 src/conf/cpu_conf.h

2 4