September 2014 - Devel - libvirt List Archives

[libvirt] [PATCH] docs: update zfs documentation
by Roman Bogorodskiy 18 Sep '14

18 Sep '14

- docs/formatstorage.html.in: document 'zfs' pool type, add it to a list of pool types that could use source physical devices - docs/storage.html.in: update a ZFS pool example XML with source physical devices, mention that starting from 1.2.9 a pool could be created from this devices by libvirt and in earlier versions user still has to create a pool manually - docs/drvbhyve.html.in: add an example with ZFS pools --- docs/drvbhyve.html.in | 16 ++++++++++++++++ docs/formatstorage.html.in | 7 ++++--- docs/storage.html.in | 18 ++++++++++++------ 3 files changed, 32 insertions(+), 9 deletions(-) diff --git a/docs/drvbhyve.html.in b/docs/drvbhyve.html.in index 71e3bc4..39afdf5 100644 --- a/docs/drvbhyve.html.in +++ b/docs/drvbhyve.html.in @@ -141,5 +141,21 @@ tweak them.</p> /usr/sbin/bhyve -c 2 -m 214 -A -I -H -P -s 0:0,hostbridge -s 3:0,virtio-net,tap0,mac=52:54:00:5d:74:e3 -s 2:0,virtio-blk,/home/user/vm1.img -s 1,lpc -l com1,/dev/nmdm0A vm1 </pre> +<h3><a name="zfsvolume">Using ZFS volumes</a></h3> + +<p>It's possible to use ZFS volumes as disk devices <span class="since">since 1.2.8</span>. +An example of domain XML device entry for that will look like:</p> + +<pre> + ... + <disk type='volume' device='disk'> + <source pool='zfspool' volume='vol1'/> + <target dev='vdb' bus='virtio'/> + </disk> + ...</pre> + +<p>Please refer to the <a href="storage.html">Storage documentation</a> for more details on storage +management.</p> + </body> </html> diff --git a/docs/formatstorage.html.in b/docs/formatstorage.html.in index e25bba7..d3e6f05 100644 --- a/docs/formatstorage.html.in +++ b/docs/formatstorage.html.in @@ -23,8 +23,9 @@ (<span class="since">since 0.7.1</span>), <code>rbd</code> (<span class="since">since 0.9.13</span>), <code>sheepdog</code> (<span class="since">since 0.10.0</span>), - or <code>gluster</code> (<span class="since">since - 1.2.0</span>). This corresponds to the + <code>gluster</code> (<span class="since">since + 1.2.0</span>) or <code>zfs</code> (<span class="since">since + 1.2.8</span>). This corresponds to the storage backend drivers listed further along in this document. </p> <h3><a name="StoragePoolFirst">General metadata</a></h3> @@ -115,7 +116,7 @@ <dt><code>device</code></dt> <dd>Provides the source for pools backed by physical devices (pool types <code>fs</code>, <code>logical</code>, <code>disk</code>, - <code>iscsi</code>). + <code>iscsi</code>, <code>zfs</code>). May be repeated multiple times depending on backend driver. Contains a single attribute <code>path</code> which is the fully qualified path to the block device node. <span class="since">Since 0.4.1</span></dd> diff --git a/docs/storage.html.in b/docs/storage.html.in index 5db79c7..3d2ffca 100644 --- a/docs/storage.html.in +++ b/docs/storage.html.in @@ -750,22 +750,28 @@ <p> This provides a pool based on the ZFS filesystem. It is currently supported on FreeBSD only. + </p> - A pool has to be created before libvirt could start using it. That - could be done using <code>zpool create</code> command. Please refer to - the ZFS documentation for details on a pool creation. - - <span class="since">Since 1.2.8</span> + <p>A pool could either be created manually using the <code>zpool create</code> + command and its name specified in the source section or <span class="since"> + since 1.2.9</span> source devices could be specified to create a pool using + libvirt. </p> + <p>Please refer to the ZFS documentation for details on a pool creation.</p> + + <p><span class="since">Since 1.2.8</span></p>. + <h3>Example pool input</h3> <pre> <pool type="zfs"> <name>myzfspool</name> <source> <name>zpoolname</name> + <device path="/dev/ada1"/> + <device path="/dev/ada2"/> </source> - </pool></pre> + </pool></pre> <h3>Valid pool format types</h3> <p> -- 2.0.2

2 2

[libvirt] [PATCH v2] storage: zfs: implement pool build and delete
by Roman Bogorodskiy 18 Sep '14

18 Sep '14

- Provide an implementation for buildPool and deletePool operations for the ZFS storage backend. - Add VIR_STORAGE_POOL_SOURCE_DEVICE flag to ZFS pool poolOptions as now we can specify devices to build pool from - storagepool.rng: add an optional 'sourceinfodev' to 'sourcezfs' and add an optional 'target' to 'poolzfs' entity - Add a couple of tests to storagepoolxml2xmltest --- docs/schemas/storagepool.rng | 6 +++ src/conf/storage_conf.c | 3 +- src/storage/storage_backend_zfs.c | 57 ++++++++++++++++++++++ tests/storagepoolxml2xmlin/pool-zfs-sourcedev.xml | 8 +++ tests/storagepoolxml2xmlin/pool-zfs.xml | 7 +++ tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml | 19 ++++++++ tests/storagepoolxml2xmlout/pool-zfs.xml | 18 +++++++ tests/storagepoolxml2xmltest.c | 2 + 8 files changed, 119 insertions(+), 1 deletion(-) create mode 100644 tests/storagepoolxml2xmlin/pool-zfs-sourcedev.xml create mode 100644 tests/storagepoolxml2xmlin/pool-zfs.xml create mode 100644 tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml create mode 100644 tests/storagepoolxml2xmlout/pool-zfs.xml diff --git a/docs/schemas/storagepool.rng b/docs/schemas/storagepool.rng index 908cc11..2d165a3 100644 --- a/docs/schemas/storagepool.rng +++ b/docs/schemas/storagepool.rng @@ -166,6 +166,9 @@ <ref name='commonmetadata'/> <ref name='sizing'/> <ref name='sourcezfs'/> + <optional> + <ref name='target'/> + </optional> </interleave> </define> @@ -386,6 +389,9 @@ <element name='source'> <interleave> <ref name='sourceinfoname'/> + <optional> + <ref name='sourceinfodev'/> + </optional> </interleave> </element> </define> diff --git a/src/conf/storage_conf.c b/src/conf/storage_conf.c index d42cde7..36696a4 100644 --- a/src/conf/storage_conf.c +++ b/src/conf/storage_conf.c @@ -282,7 +282,8 @@ static virStoragePoolTypeInfo poolTypeInfo[] = { }, {.poolType = VIR_STORAGE_POOL_ZFS, .poolOptions = { - .flags = (VIR_STORAGE_POOL_SOURCE_NAME), + .flags = (VIR_STORAGE_POOL_SOURCE_NAME | + VIR_STORAGE_POOL_SOURCE_DEVICE), .defaultFormat = VIR_STORAGE_FILE_RAW, }, }, diff --git a/src/storage/storage_backend_zfs.c b/src/storage/storage_backend_zfs.c index d8201ac..9482706 100644 --- a/src/storage/storage_backend_zfs.c +++ b/src/storage/storage_backend_zfs.c @@ -325,6 +325,61 @@ virStorageBackendZFSDeleteVol(virConnectPtr conn ATTRIBUTE_UNUSED, return ret; } +static int +virStorageBackendZFSBuildPool(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolObjPtr pool, + unsigned int flags) +{ + virCommandPtr cmd = NULL; + size_t i; + int ret = -1; + + virCheckFlags(0, -1); + + if (pool->def->source.ndevice == 0) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, + "%s", _("missing source devices")); + return -1; + } + + cmd = virCommandNewArgList(ZPOOL, "create", + pool->def->source.name, NULL); + + for (i = 0; i < pool->def->source.ndevice; i++) + virCommandAddArg(cmd, pool->def->source.devices[i].path); + + if (virCommandRun(cmd, NULL) < 0) + goto cleanup; + + ret = 0; + + cleanup: + virCommandFree(cmd); + return ret; +} + +static int +virStorageBackendZFSDeletePool(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolObjPtr pool, + unsigned int flags) +{ + virCommandPtr cmd = NULL; + int ret = -1; + + virCheckFlags(0, -1); + + cmd = virCommandNewArgList(ZPOOL, "destroy", + pool->def->source.name, NULL); + + if (virCommandRun(cmd, NULL) < 0) + goto cleanup; + + ret = 0; + + cleanup: + virCommandFree(cmd); + return ret; +} virStorageBackend virStorageBackendZFS = { .type = VIR_STORAGE_POOL_ZFS, @@ -333,6 +388,8 @@ virStorageBackend virStorageBackendZFS = { .refreshPool = virStorageBackendZFSRefreshPool, .createVol = virStorageBackendZFSCreateVol, .deleteVol = virStorageBackendZFSDeleteVol, + .buildPool = virStorageBackendZFSBuildPool, + .deletePool = virStorageBackendZFSDeletePool, .uploadVol = virStorageBackendVolUploadLocal, .downloadVol = virStorageBackendVolDownloadLocal, }; diff --git a/tests/storagepoolxml2xmlin/pool-zfs-sourcedev.xml b/tests/storagepoolxml2xmlin/pool-zfs-sourcedev.xml new file mode 100644 index 0000000..b0e0a96 --- /dev/null +++ b/tests/storagepoolxml2xmlin/pool-zfs-sourcedev.xml @@ -0,0 +1,8 @@ +<pool type="zfs"> + <name>zfs</name> + <uuid>429126d2-f4bb-45b0-b336-2e81dc6d241c</uuid> + <source> + <name>testpool</name> + <device path="/dev/ada1"/> + </source> +</pool> diff --git a/tests/storagepoolxml2xmlin/pool-zfs.xml b/tests/storagepoolxml2xmlin/pool-zfs.xml new file mode 100644 index 0000000..813342f --- /dev/null +++ b/tests/storagepoolxml2xmlin/pool-zfs.xml @@ -0,0 +1,7 @@ +<pool type="zfs"> + <name>zfs</name> + <uuid>024835f8-52b5-4226-b2b4-8c0d3afa5b2f</uuid> + <source> + <name>testpool</name> + </source> +</pool> diff --git a/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml b/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml new file mode 100644 index 0000000..bbd2e9f --- /dev/null +++ b/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml @@ -0,0 +1,19 @@ +<pool type='zfs'> + <name>zfs</name> + <uuid>429126d2-f4bb-45b0-b336-2e81dc6d241c</uuid> + <capacity unit='bytes'>0</capacity> + <allocation unit='bytes'>0</allocation> + <available unit='bytes'>0</available> + <source> + <device path='/dev/ada1'/> + <name>testpool</name> + </source> + <target> + <path>/dev/zvol/testpool</path> + <permissions> + <mode>0755</mode> + <owner>-1</owner> + <group>-1</group> + </permissions> + </target> +</pool> diff --git a/tests/storagepoolxml2xmlout/pool-zfs.xml b/tests/storagepoolxml2xmlout/pool-zfs.xml new file mode 100644 index 0000000..ff02329 --- /dev/null +++ b/tests/storagepoolxml2xmlout/pool-zfs.xml @@ -0,0 +1,18 @@ +<pool type='zfs'> + <name>zfs</name> + <uuid>024835f8-52b5-4226-b2b4-8c0d3afa5b2f</uuid> + <capacity unit='bytes'>0</capacity> + <allocation unit='bytes'>0</allocation> + <available unit='bytes'>0</available> + <source> + <name>testpool</name> + </source> + <target> + <path>/dev/zvol/testpool</path> + <permissions> + <mode>0755</mode> + <owner>-1</owner> + <group>-1</group> + </permissions> + </target> +</pool> diff --git a/tests/storagepoolxml2xmltest.c b/tests/storagepoolxml2xmltest.c index d7ae10b..8a2c0b5 100644 --- a/tests/storagepoolxml2xmltest.c +++ b/tests/storagepoolxml2xmltest.c @@ -105,6 +105,8 @@ mymain(void) DO_TEST("pool-gluster"); DO_TEST("pool-gluster-sub"); DO_TEST("pool-scsi-type-scsi-host-stable"); + DO_TEST("pool-zfs"); + DO_TEST("pool-zfs-sourcedev"); return ret == 0 ? EXIT_SUCCESS : EXIT_FAILURE; } -- 2.0.2

2 2

[libvirt] [PATCH 0/3] maint: clean up of struct XXX
by James 18 Sep '14

18 Sep '14

I find some struct XXX can be converted to typedefs. And I clean them up. James (3): maint: clean up _virDomainInterfaceStats maint: clean up _virDomainBlockStats maint: clean up _virDomainMemoryStat daemon/remote.c | 2 +- src/driver.h | 6 +++--- src/libvirt.c | 6 +++--- src/lxc/lxc_driver.c | 8 ++++---- src/openvz/openvz_driver.c | 2 +- src/qemu/qemu_driver.c | 6 +++--- src/qemu/qemu_monitor_text.c | 2 +- src/remote/remote_driver.c | 2 +- src/test/test_driver.c | 4 ++-- src/util/virstats.c | 6 +++--- src/util/virstats.h | 2 +- src/xen/block_stats.c | 4 ++-- src/xen/block_stats.h | 2 +- src/xen/xen_driver.c | 4 ++-- src/xen/xen_hypervisor.c | 4 ++-- src/xen/xen_hypervisor.h | 4 ++-- tools/virsh-domain-monitor.c | 6 +++--- 17 files changed, 35 insertions(+), 35 deletions(-) -- 1.7.12.4

3 9

[libvirt] [PATCH] Fix libvirtd crash when removing metadata
by Erik Skultety 18 Sep '14

18 Sep '14

When trying to remove nonexistent metadata from XML, libvirt daemon crashes due to dereferencing NULL pointer. Resolves https://bugzilla.redhat.com/show_bug.cgi?id=1143955 --- src/util/virxml.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/src/util/virxml.c b/src/util/virxml.c index a91da05..f386956 100644 --- a/src/util/virxml.c +++ b/src/util/virxml.c @@ -972,7 +972,9 @@ xmlNodePtr virXMLFindChildNodeByNs(xmlNodePtr root, const char *uri) { - xmlNodePtr next; + xmlNodePtr next = NULL; + if (!root) + goto cleanup; for (next = root->children; next; next = next->next) { if (next->ns && @@ -980,6 +982,7 @@ virXMLFindChildNodeByNs(xmlNodePtr root, return next; } + cleanup: return NULL; } -- 1.9.3

2 1

[libvirt] [PATCH 0/4] Memory leak fixes
by Ján Tomko 18 Sep '14

18 Sep '14

Ján Tomko (4): Fix leak in x86UpdateHostModel Fixes for domains with no iothreads audit: remove redundant NULL assignment audit: fix memory leak without WITH_AUDIT src/cpu/cpu_x86.c | 4 +++- src/qemu/qemu_cgroup.c | 2 +- src/qemu/qemu_process.c | 6 ++++-- src/util/viraudit.c | 13 +++---------- 4 files changed, 11 insertions(+), 14 deletions(-) -- 1.8.5.5

2 8

[libvirt] [PATCH] qemu: Don't fail qemuProcessAttach for IOThreads if no JSON
by John Ferlan 18 Sep '14

18 Sep '14

While doing some investigation for another bug I found that I could not qemu-attach to the process and got the following: error: Operation not supported: JSON monitor is required while running thru qemuProcessAttach. Since we can only get the data using the JSON parser and if the guest to be attached to doesn't have it we shouldn't just fail. See example in virsh qemu-attach for sample command that failed. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- I also considered removing the call from qemuProcessAttach rather than this approach. src/qemu/qemu_monitor.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 8927dbb..4342088 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -4112,11 +4112,9 @@ qemuMonitorGetIOThreads(qemuMonitorPtr mon, return -1; } - if (!mon->json) { - virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", - _("JSON monitor is required")); - return -1; - } + /* Requires JSON to make the query */ + if (!mon->json) + return 0; return qemuMonitorJSONGetIOThreads(mon, iothreads); } -- 1.9.3

2 3

[libvirt] [PATCH 1/3] bhyve: tests: fix build
by Roman Bogorodskiy 18 Sep '14

18 Sep '14

Commit b20d39a introduced a new argument for the virNetDevTapCreateInBridgePort function, however, its mock in bhyve tests wasn't updated, so the build failed. Fix build by adding this new argument to the mock version. --- tests/bhyvexml2argvmock.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tests/bhyvexml2argvmock.c b/tests/bhyvexml2argvmock.c index fa2f14b..0cbea29 100644 --- a/tests/bhyvexml2argvmock.c +++ b/tests/bhyvexml2argvmock.c @@ -22,6 +22,7 @@ int virNetDevTapCreateInBridgePort(const char *brname ATTRIBUTE_UNUSED, char **ifname, const virMacAddr *macaddr ATTRIBUTE_UNUSED, const unsigned char *vmuuid ATTRIBUTE_UNUSED, + const char *tunpath ATTRIBUTE_UNUSED, int *tapfd ATTRIBUTE_UNUSED, int tapfdSize ATTRIBUTE_UNUSED, virNetDevVPortProfilePtr virtPortProfile ATTRIBUTE_UNUSED, -- 2.1.0

2 4

[libvirt] [PATCH] qemu: fix crash with shared disks
by Ján Tomko 18 Sep '14

18 Sep '14

Commit f36a94f introduced a double free on all success paths in qemuSharedDeviceEntryInsert. Only call qemuSharedDeviceEntryFree on the error path and set entry to NULL before jumping there if the entry already is in the hash table. https://bugzilla.redhat.com/show_bug.cgi?id=1142722 --- src/qemu/qemu_conf.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index ac10b64..adc6caf 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -1011,38 +1011,36 @@ qemuSharedDeviceEntryInsert(virQEMUDriverPtr driver, const char *name) { qemuSharedDeviceEntry *entry = NULL; - int ret = -1; if ((entry = virHashLookup(driver->sharedDevices, key))) { /* Nothing to do if the shared scsi host device is already * recorded in the table. */ - if (qemuSharedDeviceEntryDomainExists(entry, name, NULL)) { - ret = 0; - goto cleanup; + if (!qemuSharedDeviceEntryDomainExists(entry, name, NULL)) { + if (VIR_EXPAND_N(entry->domains, entry->ref, 1) < 0 || + VIR_STRDUP(entry->domains[entry->ref - 1], name) < 0) { + /* entry is owned by the hash table here */ + entry = NULL; + goto error; + } } - - if (VIR_EXPAND_N(entry->domains, entry->ref, 1) < 0 || - VIR_STRDUP(entry->domains[entry->ref - 1], name) < 0) - goto cleanup; } else { if (VIR_ALLOC(entry) < 0 || VIR_ALLOC_N(entry->domains, 1) < 0 || VIR_STRDUP(entry->domains[0], name) < 0) - goto cleanup; + goto error; entry->ref = 1; if (virHashAddEntry(driver->sharedDevices, key, entry)) - goto cleanup; + goto error; } - ret = 0; + return 0; - cleanup: + error: qemuSharedDeviceEntryFree(entry, NULL); - - return ret; + return -1; } -- 1.8.5.5

2 2

[libvirt] [PATCH 0/6] Improve backing store error reporting
by Peter Krempa 17 Sep '14

17 Sep '14

Peter Krempa (6): util: Add function to check if a virStorageSource is "empty" qemu: Drop unused formatting of uuid util: storage: Allow metadata crawler to report useful errors qemu: Report better errors from broken backing chains storage: Improve error message when traversing backing chains qemu: Improve check for local storage src/libvirt_private.syms | 1 + src/qemu/qemu_domain.c | 38 ++++++-------------------------------- src/qemu/qemu_process.c | 5 ++--- src/security/virt-aa-helper.c | 2 +- src/storage/storage_driver.c | 40 +++++++++++++++++++++++++++++----------- src/storage/storage_driver.h | 3 ++- src/util/virstoragefile.c | 20 ++++++++++++++++++++ src/util/virstoragefile.h | 1 + tests/virstoragetest.c | 2 +- 9 files changed, 63 insertions(+), 49 deletions(-) -- 2.1.0

3 16

[libvirt] [PATCH] CVE-2014-3633: qemu: blkiotune: Use correct definition when looking up disk
by Peter Krempa 17 Sep '14

17 Sep '14

Live definition was used to look up the disk index while persistent one was indexed leading to a crash in qemuDomainGetBlockIoTune. Use the correct def and report a nice error. Unfortunately it's accessible via read-only connection. Introduced in: eca96694a7f992be633d48d5ca03cedc9bbc3c9aa (v0.9.8) Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1140724 Reported-by: Luyao Huang <lhuang(a)redhat.com> Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- src/qemu/qemu_driver.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index a5a49ac..209c40e 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -16317,9 +16317,13 @@ qemuDomainGetBlockIoTune(virDomainPtr dom, } if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - int idx = virDomainDiskIndexByName(vm->def, disk, true); - if (idx < 0) + int idx = virDomainDiskIndexByName(persistentDef, disk, true); + if (idx < 0) { + virReportError(VIR_ERR_INVALID_ARG, + _("disk '%s' was not found in the domain config"), + disk); goto endjob; + } reply = persistentDef->disks[idx]->blkdeviotune; } -- 2.1.0

2 2