September 2011 - Devel - libvirt List Archives

[libvirt] [PATCH] selinux: Correctly report warning if virt_use_nfs not set
by Michal Privoznik 23 Sep '11

23 Sep '11

Previous patch c9b37fee tried to deal with virt_use_nfs. But setfilecon() returns EOPNOTSUPP on NFS so we need to move the warning to else branch. --- src/security/security_selinux.c | 27 +++++++++++++++------------ 1 files changed, 15 insertions(+), 12 deletions(-) diff --git a/src/security/security_selinux.c b/src/security/security_selinux.c index 028f5b2..9a9a305 100644 --- a/src/security/security_selinux.c +++ b/src/security/security_selinux.c @@ -420,23 +420,26 @@ SELinuxSetFilecon(const char *path, char *tcon) * virt_use_{nfs,usb,pci} boolean tunables to allow it... */ if (setfilecon_errno != EOPNOTSUPP) { - const char *errmsg; - if ((virStorageFileIsSharedFSType(path, - VIR_STORAGE_FILE_SHFS_NFS) == 1) && - security_get_boolean_active("virt_use_nfs") != 1) { - errmsg = _("unable to set security context '%s' on '%s'. " - "Consider setting virt_use_nfs"); - } else { - errmsg = _("unable to set security context '%s' on '%s'"); - } virReportSystemError(setfilecon_errno, - errmsg, + _("unable to set security context '%s' on '%s'"), tcon, path); if (security_getenforce() == 1) return -1; } else { - VIR_INFO("Setting security context '%s' on '%s' not supported", - tcon, path); + const char *msg; + if ((virStorageFileIsSharedFSType(path, + VIR_STORAGE_FILE_SHFS_NFS) == 1) && + security_get_boolean_active("virt_use_nfs") != 1) { + msg = _("Setting security context '%s' on '%s' not supported. " + "Consider setting virt_use_nfs"); + if (security_getenforce() == 1) + VIR_WARN(msg, tcon, path); + else + VIR_INFO(msg, tcon, path); + } else { + VIR_INFO(_("Setting security context '%s' " + "on '%s' not supported"), tcon, path); + } } } return 0; -- 1.7.3.4

3 3

[libvirt] [PATCH] storage: Do not use comma as seperator for lvs output
by Osier Yang 23 Sep '11

23 Sep '11

* src/storage/storage_backend_logical.c: If a logical vol is created with multiple stripes. (e.g. --stripes 3), the "device" field of lvs output will have multiple fileds which are seperated by comma. It means the RE we write in the codes will not work well anymore. E.g. (lvs output for a stripped vol, uses "#" as seperator here): test_stripes##fSLSZH-zAS2-yAIb-n4mV-Al9u-HA3V-oo9K1B#\ /dev/sdc1(10240),/dev/sdd1(0)#42949672960#4194304 The RE we uses: const char *regexes[] = { "^\\s*(\\S+),(\\S*),(\\S+),(\\S+)\$(\\S+)\$,(\\S+),([0-9]+),?\\s*$" }; This patch changes the seperator into "#" to fix the problem. Related RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=727474 --- src/storage/storage_backend_logical.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/storage/storage_backend_logical.c b/src/storage/storage_backend_logical.c index 4f42047..6c13a6b 100644 --- a/src/storage/storage_backend_logical.c +++ b/src/storage/storage_backend_logical.c @@ -199,7 +199,7 @@ virStorageBackendLogicalFindLVs(virStoragePoolObjPtr pool, 7 }; const char *prog[] = { - LVS, "--separator", ",", "--noheadings", "--units", "b", + LVS, "--separator", "#", "--noheadings", "--units", "b", "--unbuffered", "--nosuffix", "--options", "lv_name,origin,uuid,devices,seg_size,vg_extent_size", pool->def->source.name, NULL -- 1.7.6

5 10

Re: [libvirt] [PATCH] qemu: Allow domain reboot after core dump
by Sheldon 23 Sep '11

23 Sep '11

yes, the patch can work, when I execute virsh dump --reboot <domain> <file> and the type of file is: QEMU's suspend to disk image but the guest does not reboot. The guest does nothing. but I try virsh shutdown <domain>, the guest Ubuntu OS pop up a dialog box with four options: Shut Down, Restart, Suspend and HIbernate. -------- ???? -------- > ??: [libvirt] [PATCH] qemu: Allow domain reboot after core dump > ??: Tue, 20 Sep 2011 13:34:27 +0200 > ???: Michal Privoznik <mprivozn(a)redhat.com> > ???: libvir-list(a)redhat.com > > > > This patch introduces possibility to reboot domain after core dump > finishes. The new flag VIR_DUMP_REBOOT was added to > virDomainCoreDumpFlags. The new functionality is accessible via virsh > too: virsh dump --reboot<domain> > --- > include/libvirt/libvirt.h.in | 1 + > src/qemu/qemu_driver.c | 8 +++++++- > tools/virsh.c | 3 +++ > 3 files changed, 11 insertions(+), 1 deletions(-) > > diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in > index 39155a6..8c41f5a 100644 > --- a/include/libvirt/libvirt.h.in > +++ b/include/libvirt/libvirt.h.in > @@ -748,6 +748,7 @@ typedef enum { > VIR_DUMP_CRASH = (1<< 0), /* crash after dump */ > VIR_DUMP_LIVE = (1<< 1), /* live dump */ > VIR_DUMP_BYPASS_CACHE = (1<< 2), /* avoid file system cache pollution */ > + VIR_DUMP_REBOOT = (1<< 3), /* reboot domain after dump finishes */ > } virDomainCoreDumpFlags; > > /* Domain migration flags. */ > diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c > index e2f428f..22576a8 100644 > --- a/src/qemu/qemu_driver.c > +++ b/src/qemu/qemu_driver.c > @@ -3104,7 +3104,8 @@ static int qemudDomainCoreDump(virDomainPtr dom, > int ret = -1; > virDomainEventPtr event = NULL; > > - virCheckFlags(VIR_DUMP_LIVE | VIR_DUMP_CRASH | VIR_DUMP_BYPASS_CACHE, -1); > + virCheckFlags(VIR_DUMP_LIVE | VIR_DUMP_CRASH | > + VIR_DUMP_BYPASS_CACHE | VIR_DUMP_REBOOT, -1); > > qemuDriverLock(driver); > vm = virDomainFindByUUID(&driver->domains, dom->uuid); > @@ -3189,6 +3190,11 @@ cleanup: > if (event) > qemuDomainEventQueue(driver, event); > qemuDriverUnlock(driver); > + > + if ((ret == 0)&& (flags& VIR_DUMP_REBOOT)) { > + qemuDomainReboot(dom, 0); > + } > + > return ret; > } > > diff --git a/tools/virsh.c b/tools/virsh.c > index d575425..74f6a79 100644 > --- a/tools/virsh.c > +++ b/tools/virsh.c > @@ -2840,6 +2840,7 @@ static const vshCmdOptDef opts_dump[] = { > {"crash", VSH_OT_BOOL, 0, N_("crash the domain after core dump")}, > {"bypass-cache", VSH_OT_BOOL, 0, > N_("avoid file system cache when saving")}, > + {"reboot", VSH_OT_BOOL, 0, N_("reboot the domain after core dump")}, > {"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or uuid")}, > {"file", VSH_OT_DATA, VSH_OFLAG_REQ, N_("where to dump the core")}, > {NULL, 0, 0, NULL} > @@ -2869,6 +2870,8 @@ cmdDump(vshControl *ctl, const vshCmd *cmd) > flags |= VIR_DUMP_CRASH; > if (vshCommandOptBool(cmd, "bypass-cache")) > flags |= VIR_DUMP_BYPASS_CACHE; > + if (vshCommandOptBool(cmd, "reboot")) > + flags |= VIR_DUMP_REBOOT; > > if (virDomainCoreDump(dom, to, flags)< 0) { > vshError(ctl, _("Failed to core dump domain %s to %s"), name, to); > -- > 1.7.3.4 > > -- > libvir-list mailing list > libvir-list(a)redhat.com > https://www.redhat.com/mailman/listinfo/libvir-list >

1 0

Re: [libvirt] [PATCH] qemu: Allow domain reboot after core dump
by Sheldon 23 Sep '11

23 Sep '11

yes, the patch can work, when I execute virsh dump --reboot <domian> <file> and the type of file is: QEMU's suspend to disk image > -------- ???? -------- > ??: [libvirt] [PATCH] qemu: Allow domain reboot after core dump > ??: Tue, 20 Sep 2011 13:34:27 +0200 > ???: Michal Privoznik <mprivozn(a)redhat.com> > ???: libvir-list(a)redhat.com > > > > This patch introduces possibility to reboot domain after core dump > finishes. The new flag VIR_DUMP_REBOOT was added to > virDomainCoreDumpFlags. The new functionality is accessible via virsh > too: virsh dump --reboot<domain> > --- > include/libvirt/libvirt.h.in | 1 + > src/qemu/qemu_driver.c | 8 +++++++- > tools/virsh.c | 3 +++ > 3 files changed, 11 insertions(+), 1 deletions(-) > > diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in > index 39155a6..8c41f5a 100644 > --- a/include/libvirt/libvirt.h.in > +++ b/include/libvirt/libvirt.h.in > @@ -748,6 +748,7 @@ typedef enum { > VIR_DUMP_CRASH = (1<< 0), /* crash after dump */ > VIR_DUMP_LIVE = (1<< 1), /* live dump */ > VIR_DUMP_BYPASS_CACHE = (1<< 2), /* avoid file system cache pollution */ > + VIR_DUMP_REBOOT = (1<< 3), /* reboot domain after dump finishes */ > } virDomainCoreDumpFlags; > > /* Domain migration flags. */ > diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c > index e2f428f..22576a8 100644 > --- a/src/qemu/qemu_driver.c > +++ b/src/qemu/qemu_driver.c > @@ -3104,7 +3104,8 @@ static int qemudDomainCoreDump(virDomainPtr dom, > int ret = -1; > virDomainEventPtr event = NULL; > > - virCheckFlags(VIR_DUMP_LIVE | VIR_DUMP_CRASH | VIR_DUMP_BYPASS_CACHE, -1); > + virCheckFlags(VIR_DUMP_LIVE | VIR_DUMP_CRASH | > + VIR_DUMP_BYPASS_CACHE | VIR_DUMP_REBOOT, -1); > > qemuDriverLock(driver); > vm = virDomainFindByUUID(&driver->domains, dom->uuid); > @@ -3189,6 +3190,11 @@ cleanup: > if (event) > qemuDomainEventQueue(driver, event); > qemuDriverUnlock(driver); > + > + if ((ret == 0)&& (flags& VIR_DUMP_REBOOT)) { > + qemuDomainReboot(dom, 0); > + } > + > return ret; > } > > diff --git a/tools/virsh.c b/tools/virsh.c > index d575425..74f6a79 100644 > --- a/tools/virsh.c > +++ b/tools/virsh.c > @@ -2840,6 +2840,7 @@ static const vshCmdOptDef opts_dump[] = { > {"crash", VSH_OT_BOOL, 0, N_("crash the domain after core dump")}, > {"bypass-cache", VSH_OT_BOOL, 0, > N_("avoid file system cache when saving")}, > + {"reboot", VSH_OT_BOOL, 0, N_("reboot the domain after core dump")}, > {"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or uuid")}, > {"file", VSH_OT_DATA, VSH_OFLAG_REQ, N_("where to dump the core")}, > {NULL, 0, 0, NULL} > @@ -2869,6 +2870,8 @@ cmdDump(vshControl *ctl, const vshCmd *cmd) > flags |= VIR_DUMP_CRASH; > if (vshCommandOptBool(cmd, "bypass-cache")) > flags |= VIR_DUMP_BYPASS_CACHE; > + if (vshCommandOptBool(cmd, "reboot")) > + flags |= VIR_DUMP_REBOOT; > > if (virDomainCoreDump(dom, to, flags)< 0) { > vshError(ctl, _("Failed to core dump domain %s to %s"), name, to); > -- > 1.7.3.4 > > -- > libvir-list mailing list > libvir-list(a)redhat.com > https://www.redhat.com/mailman/listinfo/libvir-list >

1 0

[libvirt] [PATCH] virsh: Improve virsh manual for virsh memtune command
by Peter Krempa 23 Sep '11

23 Sep '11

Commit 0a22f54 added --min-guarantee option for the memtune command. This option is supported only by the ESX hypervisor. This patch adds a statement about this fact, to prevent user confusion. This patch also adds explanation how to clear/set to unlimited the memory tunables. (documments the -1 value). --- tools/virsh.pod | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index 88dee26..7f46ee4 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -959,6 +959,7 @@ Allows you to display or set the domain memory parameters. Without flags, the current settings are displayed; with a flag, the appropriate limit is adjusted if supported by the hypervisor. LXC and QEMU/KVM support I<--hard-limit>, I<--soft-limit>, and I<--swap-hard-limit>. +I<--min-guarantee> is supported only by ESX hypervisor yet. If I<--live> is specified, affect a running guest. If I<--config> is specified, affect the next boot of a persistent guest. @@ -997,6 +998,8 @@ value are kilobytes (i.e. blocks of 1024 bytes). =back +Specifying -1 as a value for these limits is interpreted as unlimited. + =item B<blkiotune> I<domain-id> [I<--weight> B<weight>] [[I<--config>] [I<--live>] | [I<--current>]] -- 1.7.3.4

2 2

[libvirt] Release of libvirt-0.9.6
by Daniel Veillard 23 Sep '11

23 Sep '11

That's what is now usually called a brown paper bag release (dixit Linus) [*] since 0.9.5 had broken shudown with the two last released versions of QEmu 0.9.6 should be available at the usual place: ftp://libvirt.org/libvirt/ and obviously only contains a few bug fixes and local improvements. People should not ship 0.9.5, and use 0.9.6 instead, thanks in advance ! Bug Fixes: - snapshot: fix logic bug in qemu undefine (Eric Blake) - sanlock: fix memory leak (Eric Blake) - virsh: fix regression in argv parsing (Eric Blake) - storage: Wait udev events are handled before removing lvm vol (Osier Yang) - qemu: avoid dereferencing a NULL pointer (Alex Jia) - qemu: Avoid loop of fake reboots (Jiri Denemark) - qemu: Fix shutdown regression with buggy qemu (Jiri Denemark) - remote: fix crash on OOM (Eric Blake) Improvements: - snapshot: also delete empty directory (Eric Blake) - snapshot: remove snapshot metadata on transient exit (Eric Blake) - snapshot: prepare to remove transient snapshot metadata (Eric Blake) - virsh: More friendly err if no pool is specified for looking up a vol (Osier Yang) - tests: improve test failure diagnosis (Oskari Saarenmaa) - API: prefer to use NULLSTR macro (Osier Yang) - xen: use typical allocations (Eric Blake) - virsh: Remove useless codes of cmdVolPath (Osier Yang) Portability: - build: silence warning on 32-bit build (Eric Blake) Thanks to all who reported and helped fix those issues quickly !!! Daniel [*] and no I won't carry one, they are non-existent where I live for context http://lwn.net/1999/0204/kernel.php3 -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 1

[libvirt] [PATCH 0/9] Implement keepalive protocol for libvirt RPC
by Jiri Denemark 22 Sep '11

22 Sep '11

This allows us to detect broken connections between server and client without waiting for TCP timeout and dead deamon/client. By default a connection is considered broken after about 30 seconds of no messages received from remote party. After that period, the connection is automatically closed. The main reason for implementing this is that peer-to-peer migration can now be canceled when a connection between source and target breaks. Although this will really work only after qemu fixes migrate_cancel command so that it doesn't block when outgoing TCP buffers are full. Jiri Denemark (9): Define keepalive protocol Implement common keepalive handling Introduce two public APIs for keepalive protocol Implement keepalive protocol in libvirt daemon Add support for non-blocking calls in client RPC Add support for async close of client RPC socket Implement keepalive protocol in remote driver Add keepalive support into domain-events examples qemu: Add support for keepalive messages during p2p migration .gitignore | 1 + daemon/libvirtd.aug | 4 + daemon/libvirtd.c | 11 + daemon/libvirtd.conf | 15 + daemon/remote.c | 38 ++ examples/domain-events/events-c/event-test.c | 11 + examples/domain-events/events-python/event-test.py | 3 + include/libvirt/libvirt.h.in | 5 + po/POTFILES.in | 1 + src/Makefile.am | 13 +- src/driver.h | 9 + src/libvirt.c | 107 +++++ src/libvirt_internal.h | 10 +- src/libvirt_public.syms | 6 + src/qemu/libvirtd_qemu.aug | 2 + src/qemu/qemu.conf | 16 + src/qemu/qemu_conf.c | 11 + src/qemu/qemu_conf.h | 3 + src/qemu/qemu_migration.c | 10 + src/qemu/test_libvirtd_qemu.aug | 6 + src/remote/remote_driver.c | 30 ++ src/remote/remote_protocol.x | 2 +- src/rpc/virkeepalive.c | 464 ++++++++++++++++++++ src/rpc/virkeepalive.h | 58 +++ src/rpc/virkeepaliveprotocol.x | 8 + src/rpc/virnetclient.c | 321 ++++++++++++-- src/rpc/virnetclient.h | 5 + src/rpc/virnetserver.c | 10 + src/rpc/virnetserver.h | 2 + src/rpc/virnetserverclient.c | 126 +++++- src/rpc/virnetserverclient.h | 6 + 31 files changed, 1254 insertions(+), 60 deletions(-) create mode 100644 src/rpc/virkeepalive.c create mode 100644 src/rpc/virkeepalive.h create mode 100644 src/rpc/virkeepaliveprotocol.x -- 1.7.6.1

1 9

[libvirt] [PATCH] virsh: Do not ignore the specified flags for cmdSaveImageDefine
by Osier Yang 22 Sep '11

22 Sep '11

--- tools/virsh.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 4b9e662..02db278 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -2346,7 +2346,7 @@ cmdSaveImageDefine(vshControl *ctl, const vshCmd *cmd) if (virFileReadAll(xmlfile, 8192, &xml) < 0) goto cleanup; - if (virDomainSaveImageDefineXML(ctl->conn, file, xml, 0) < 0) { + if (virDomainSaveImageDefineXML(ctl->conn, file, xml, flags) < 0) { vshError(ctl, _("Failed to update %s"), file); goto cleanup; } -- 1.7.6

2 2

[libvirt] [PATCH] qemu: Transfer inactive XML among cookie
by Michal Privoznik 22 Sep '11

22 Sep '11

If a domain has inactive XML we want to transfer it to destination when migrating with VIR_MIGRATE_PERSIST_DEST. In order to harm the migration protocol as least as possible, a optional cookie was chosen. --- src/qemu/qemu_migration.c | 91 ++++++++++++++++++++++++++++++++++++++++---- 1 files changed, 82 insertions(+), 9 deletions(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index d9f8d93..bc98708 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -64,6 +64,7 @@ VIR_ENUM_IMPL(qemuMigrationJobPhase, QEMU_MIGRATION_PHASE_LAST, enum qemuMigrationCookieFlags { QEMU_MIGRATION_COOKIE_FLAG_GRAPHICS, QEMU_MIGRATION_COOKIE_FLAG_LOCKSTATE, + QEMU_MIGRATION_COOKIE_FLAG_PERSISTENT, QEMU_MIGRATION_COOKIE_FLAG_LAST }; @@ -71,11 +72,12 @@ enum qemuMigrationCookieFlags { VIR_ENUM_DECL(qemuMigrationCookieFlag); VIR_ENUM_IMPL(qemuMigrationCookieFlag, QEMU_MIGRATION_COOKIE_FLAG_LAST, - "graphics", "lockstate"); + "graphics", "lockstate", "persistent"); enum qemuMigrationCookieFeatures { QEMU_MIGRATION_COOKIE_GRAPHICS = (1 << QEMU_MIGRATION_COOKIE_FLAG_GRAPHICS), QEMU_MIGRATION_COOKIE_LOCKSTATE = (1 << QEMU_MIGRATION_COOKIE_FLAG_LOCKSTATE), + QEMU_MIGRATION_COOKIE_PERSISTENT = (1 << QEMU_MIGRATION_COOKIE_FLAG_PERSISTENT), }; typedef struct _qemuMigrationCookieGraphics qemuMigrationCookieGraphics; @@ -110,6 +112,9 @@ struct _qemuMigrationCookie { /* If (flags & QEMU_MIGRATION_COOKIE_GRAPHICS) */ qemuMigrationCookieGraphicsPtr graphics; + + /* If (flags & QEMU_MIGRATION_COOKIE_PERSISTENT) */ + virDomainDefPtr persistent; }; static void qemuMigrationCookieGraphicsFree(qemuMigrationCookieGraphicsPtr grap) @@ -334,6 +339,26 @@ qemuMigrationCookieAddLockstate(qemuMigrationCookiePtr mig, } +static int +qemuMigrationCookieAddPersistent(qemuMigrationCookiePtr mig, + virDomainObjPtr dom) +{ + if (mig->flags & QEMU_MIGRATION_COOKIE_PERSISTENT) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Migration persistent data already present")); + return -1; + } + + if (!dom->newDef) + return 0; + + mig->persistent = dom->newDef; + mig->flags |= QEMU_MIGRATION_COOKIE_PERSISTENT; + mig->flagsMandatory |= QEMU_MIGRATION_COOKIE_PERSISTENT; + return 0; +} + + static void qemuMigrationCookieGraphicsXMLFormat(virBufferPtr buf, qemuMigrationCookieGraphicsPtr grap) @@ -354,10 +379,12 @@ static void qemuMigrationCookieGraphicsXMLFormat(virBufferPtr buf, static void qemuMigrationCookieXMLFormat(virBufferPtr buf, - qemuMigrationCookiePtr mig) + qemuMigrationCookiePtr mig, + struct qemud_driver *driver) { char uuidstr[VIR_UUID_STRING_BUFLEN]; char hostuuidstr[VIR_UUID_STRING_BUFLEN]; + char *domXML; int i; virUUIDFormat(mig->uuid, uuidstr); @@ -388,15 +415,25 @@ static void qemuMigrationCookieXMLFormat(virBufferPtr buf, virBufferAddLit(buf, " </lockstate>\n"); } + if ((mig->flags & QEMU_MIGRATION_COOKIE_PERSISTENT) && + mig->persistent) { + domXML = qemuDomainDefFormatXML(driver, mig->persistent, + VIR_DOMAIN_XML_INACTIVE | + VIR_DOMAIN_XML_SECURE); + virBufferAdd(buf, domXML, -1); + VIR_FREE(domXML); + } + virBufferAddLit(buf, "</qemu-migration>\n"); } -static char *qemuMigrationCookieXMLFormatStr(qemuMigrationCookiePtr mig) +static char *qemuMigrationCookieXMLFormatStr(qemuMigrationCookiePtr mig, + struct qemud_driver *driver) { virBuffer buf = VIR_BUFFER_INITIALIZER; - qemuMigrationCookieXMLFormat(&buf, mig); + qemuMigrationCookieXMLFormat(&buf, mig, driver); if (virBufferError(&buf)) { virReportOOMError(); @@ -460,6 +497,8 @@ error: static int qemuMigrationCookieXMLParse(qemuMigrationCookiePtr mig, + struct qemud_driver *driver, + xmlDocPtr doc, xmlXPathContextPtr ctxt, unsigned int flags) { @@ -583,6 +622,25 @@ qemuMigrationCookieXMLParse(qemuMigrationCookiePtr mig, VIR_FREE(mig->lockState); } + if ((flags & QEMU_MIGRATION_COOKIE_PERSISTENT) && + virXPathBoolean("count(./domain) > 0", ctxt)) { + if ((n = virXPathNodeSet("./domain", ctxt, &nodes)) > 1) { + qemuReportError(VIR_ERR_INTERNAL_ERROR, + _("Too many domain elements in " + "migration cookie: %d"), + n); + goto error; + } + mig->persistent = virDomainDefParseNode(driver->caps, doc, nodes[0], + -1, VIR_DOMAIN_XML_INACTIVE); + if (!mig->persistent) { + /* virDomainDefParseNode already reported + * an error for us */ + goto error; + } + VIR_FREE(nodes); + } + return 0; error: @@ -594,6 +652,7 @@ error: static int qemuMigrationCookieXMLParseStr(qemuMigrationCookiePtr mig, + struct qemud_driver *driver, const char *xml, unsigned int flags) { @@ -606,7 +665,7 @@ qemuMigrationCookieXMLParseStr(qemuMigrationCookiePtr mig, if (!(doc = virXMLParseStringCtxt(xml, _("(qemu_migration_cookie)"), &ctxt))) goto cleanup; - ret = qemuMigrationCookieXMLParse(mig, ctxt, flags); + ret = qemuMigrationCookieXMLParse(mig, driver, doc, ctxt, flags); cleanup: xmlXPathFreeContext(ctxt); @@ -637,7 +696,11 @@ qemuMigrationBakeCookie(qemuMigrationCookiePtr mig, qemuMigrationCookieAddLockstate(mig, driver, dom) < 0) return -1; - if (!(*cookieout = qemuMigrationCookieXMLFormatStr(mig))) + if (flags & QEMU_MIGRATION_COOKIE_PERSISTENT && + qemuMigrationCookieAddPersistent(mig, dom) < 0) + return -1; + + if (!(*cookieout = qemuMigrationCookieXMLFormatStr(mig, driver))) return -1; *cookieoutlen = strlen(*cookieout) + 1; @@ -672,6 +735,7 @@ qemuMigrationEatCookie(struct qemud_driver *driver, if (cookiein && cookieinlen && qemuMigrationCookieXMLParseStr(mig, + driver, cookiein, flags) < 0) goto error; @@ -1575,7 +1639,8 @@ cleanup: } if (ret == 0 && - qemuMigrationBakeCookie(mig, driver, vm, cookieout, cookieoutlen, 0) < 0) + qemuMigrationBakeCookie(mig, driver, vm, cookieout, cookieoutlen, + QEMU_MIGRATION_COOKIE_PERSISTENT ) < 0) VIR_WARN("Unable to encode migration cookie"); qemuMigrationCookieFree(mig); @@ -2477,6 +2542,7 @@ qemuMigrationFinish(struct qemud_driver *driver, int newVM = 1; qemuMigrationCookiePtr mig = NULL; virErrorPtr orig_err = NULL; + int cookie_flags = 0; VIR_DEBUG("driver=%p, dconn=%p, vm=%p, cookiein=%s, cookieinlen=%d, " "cookieout=%p, cookieoutlen=%p, flags=%lx, retcode=%d", @@ -2490,7 +2556,11 @@ qemuMigrationFinish(struct qemud_driver *driver, v3proto ? QEMU_MIGRATION_PHASE_FINISH3 : QEMU_MIGRATION_PHASE_FINISH2); - if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, cookieinlen, 0))) + if (flags & VIR_MIGRATE_PERSIST_DEST) + cookie_flags |= QEMU_MIGRATION_COOKIE_PERSISTENT; + + if (!(mig = qemuMigrationEatCookie(driver, vm, cookiein, + cookieinlen, cookie_flags))) goto endjob; /* Did the migration go as planned? If yes, return the domain @@ -2510,7 +2580,10 @@ qemuMigrationFinish(struct qemud_driver *driver, if (vm->persistent) newVM = 0; vm->persistent = 1; - vmdef = virDomainObjGetPersistentDef(driver->caps, vm); + if (mig->persistent) + vm->newDef = vmdef = mig->persistent; + else + vmdef = virDomainObjGetPersistentDef(driver->caps, vm); if (virDomainSaveConfig(driver->configDir, vmdef) < 0) { /* Hmpf. Migration was successful, but making it persistent * was not. If we report successful, then when this domain -- 1.7.3.4

3 10

[libvirt] [PATCH 0/3] snapshot: auto-cleanup metadata of transient domains
by Eric Blake 22 Sep '11

22 Sep '11

I intended to do this ever since I changed the API of virDomainUndefine to reject undefine of persistent guests with snapshots, but am only now getting around to it. This missed 0.9.5, but I think it is worth including in 0.9.6 even if we have a short release cycle to work around the qemu shutdown issue. Eric Blake (3): snapshot: fix logic bug in qemu undefine snapshot: prepare to remove transient snapshot metadata snapshot: remove snapshot metadata on transient exit src/conf/domain_conf.c | 4 +- src/qemu/qemu_domain.c | 260 ++++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_domain.h | 36 ++++++ src/qemu/qemu_driver.c | 291 +++----------------------------------------- src/qemu/qemu_migration.c | 8 +- src/qemu/qemu_process.c | 10 +- 6 files changed, 327 insertions(+), 282 deletions(-) -- 1.7.4.4

2 6