Commit Graph

15663 Commits

Author SHA1 Message Date
Jim Fehlig
cfad607b23 libxl: handle on_crash coredump actions
Add support for coredump-{destroy,restart} actions of <on_crash> event.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-24 10:39:44 -07:00
Jim Fehlig
c2de456e4e libxl: add dump dir to libxlDriverConfig object
Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-24 10:27:53 -07:00
Jim Fehlig
51b9b39127 libxl: honor domain lifecycle event configuration
The libxl driver was ignoring the <on_*> domain event configuration,
causing e.g. a domain to be rebooted even when on_reboot is set to
destroy.

This patch honors the <on_*> configuration in the shutdown event
handler.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-24 10:26:52 -07:00
Richard Weinberger
6fb42d7cdc Ensure systemd cgroup ownership is delegated to container with userns
This function is needed for user namespaces, where we need to chmod()
the cgroup to the initial uid/gid such that systemd is allowed to
use the cgroup.

Signed-off-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-24 15:35:47 +00:00
Roman Bogorodskiy
8ca5f46c59 bhyve: implement node information reporting
- Implement nodeGetCPUStats using nodeGetCPUStats()
- Implement nodeGetMemoryStats using nodeGetMemoryStats()
2014-02-24 19:03:46 +04:00
Daniel P. Berrange
66e3a3e914 Add virStringReplace method for substring replacement
Add a virStringReplace method to virstring.{h,c} to perform
substring matching and replacement

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-24 10:51:22 +00:00
Manuel VIVES
12aa71dfde Add virStringSearch method for regex matching
Add a virStringSearch method to virstring.{c,h} which performs
a regex match against a string and returns the matching substrings.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-24 10:46:28 +00:00
Michal Privoznik
68954fb25c virNetServerRun: Notify systemd that we're accepting clients
Systemd does not forget about the cases, where client service needs to
wait for daemon service to initialize and start accepting new clients.
Setting a dependency in client is not enough as systemd doesn't know
when the daemon has initialized itself and started accepting new
clients. However, it offers a mechanism to solve this. The daemon needs
to call a special systemd function by which the daemon tells "I'm ready
to accept new clients". This is exactly what we need with
libvirtd-guests (client) and libvirtd (daemon). So now, with this
change, libvirt-guests.service is invoked not any sooner than
libvirtd.service calls the systemd notify function.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-02-24 10:54:48 +01:00
Michal Privoznik
4e7fc8305a libvirt-guests: Wait for libvirtd to initialize
I've noticed that in some cases systemd was quick enough and even
if libvirt-guests.service is marked to be started after the
libvirtd.service my guests were not resumed as
libvirt-guests.sh failed to connect. This is because of a
simple fact: systemd correctly starts libvirt-guests after it
execs libvirtd. However, the daemon is not able to accept
connections right from the start. It's doing some
initialization which may take ages. This problem is not limited
to systemd only, indeed. Any init system that is able to startup
services in parallel (e.g. OpenRC) may run into this situation.
The fix is to try connecting not only once, but continuously a few
times with a small sleep in between tries.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-02-24 10:54:37 +01:00
Michal Privoznik
ba79e3879e virSystemdCreateMachine: Set dependencies for slices
https://bugzilla.redhat.com/show_bug.cgi?id=1031696

When creating a new domain, we let systemd know about it by calling
CreateMachine() function via dbus. Systemd then creates a scope and
places domain into it. However, later when the host is shutting
down, systemd computes the shutdown order to see what processes can
be shut down in parallel. And since we were not setting
dependencies at all, the slices (and thus domains) were most likely
killed before libvirt-guests.service. So user domains that had to
be saved, shut off, whatever were in fact killed.  This problem can
be solved by letting systemd know that scopes we're creating must
not be killed before libvirt-guests.service.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-02-24 10:21:00 +01:00
Eric Blake
e203bef592 build: fix build on 32-bit hosts
vircgrouptest.c: In function 'testCgroupGetPercpuStats':
vircgrouptest.c:543: warning: integer constatnt is too large for 'long' type

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-21 17:21:58 -07:00
Eric Blake
1cf11064f1 maint: update to latest gnulib, for older autoconf
Based on a report from Pavel Hrdina, gnulib was fixed to support
AC_PROG_SED even when using ancient autoconf 2.59 of RHEL 5.

* .gnulib: Update to latest, to fix build on RHEL 5.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-21 17:18:34 -07:00
Ján Tomko
57e17a74b7 Ignore additional fields in iscsiadm output
There has been a new field introduced in iscsiadm --mode session
output [1], but our regex only expects four fields. This breaks
startup of iscsi pools:
error: Failed to start pool iscsi
error: internal error: cannot find session

Fix this by ignoring anything after the fourth field.

https://bugzilla.redhat.com/show_bug.cgi?id=1067173

[1] https://github.com/mikechristie/open-iscsi/commit/181af9a
2014-02-21 10:35:57 +01:00
Ján Tomko
abf1daf0d7 Add a stub for virCgroupGetDomainTotalCpuStats
Commit 6515889 broke the build on FreeBSD:
In function `qemuDomainGetCPUStats':
/../../src/qemu/qemu_driver.c:16102:
undefined reference to `virCgroupGetDomainTotalCpuStats'
2014-02-21 09:10:48 +01:00
Eric Blake
96d947ce03 maint: update to latest gnulib
Among other things, gnulib now allows overriding of sed during
'make syntax-check'.

* .gnulib: Update to latest, for maint.mk improvements.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 16:46:47 -07:00
Eric Blake
de87691ff0 virsh: add net-event command
Add 'virsh net-event --list' and 'virsh net-event [net] --event=name
[--loop] [--timeout]'.  Very similar to 'virsh event'.

* tools/virsh.pod (net-event): Document new command.
* tools/virsh-network.c (vshNetworkEventToString, vshNetEventData)
(vshEventLifecyclePrint, cmdNetworkEvent): New struct and
functions.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 16:23:27 -07:00
Eric Blake
99fa96c390 virsh: add event command, for lifecycle events
Add 'virsh event --list' and 'virsh event [dom] --event=name
[--loop] [--timeout]'.  Borrows somewhat from event-test.c,
but defaults to a one-shot notification, and takes advantage
of the event loop integration to allow Ctrl-C to interrupt the
wait for an event.  For now, this just does lifecycle events.

* tools/virsh.pod (event): Document new command.
* tools/virsh-domain.c (vshDomainEventToString)
(vshDomainEventDetailToString, vshDomEventData)
(vshEventLifecyclePrint, cmdEvent): New struct and functions.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 16:23:27 -07:00
Eric Blake
4c29530d8e virsh: common code for waiting for an event
I plan to add 'virsh event' to virsh-domain.c and 'virsh
net-event' to virsh-network.c; but as they will share quite
a bit of common boilerplate, it's better to set that up now
in virsh.c.

* tools/virsh.h (_vshControl): Add fields.
(vshEventStart, vshEventWait, vshEventDone, vshEventCleanup): New
prototypes.
* tools/virsh.c (vshEventFd, vshEventOldAction, vshEventInt)
(vshEventTimeout): New helper variables and functions.
(vshEventStart, vshEventWait, vshEventDone, vshEventCleanup):
Implement new functions.
(vshInit, vshDeinit, main): Manage event timeout.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 16:23:27 -07:00
Eric Blake
5093b047ea virsh: common code for parsing --seconds
Several virsh commands ask for a --timeout parameter in
seconds, then use it to control interfaces that operate on
millisecond limits; I also plan on adding a 'virsh event'
command that also does this.  Factor this into a common
function.

* tools/virsh.h (vshCommandOptTimeoutToMs): New prototype.
* tools/virsh.c (vshCommandOptTimeoutToMs): New function.
* tools/virsh-domain.c (cmdBlockCommit, cmdBlockCopy)
(cmdBlockPull, cmdMigrate): Use it.
(vshWatchJob): Adjust timeout scale.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 16:23:07 -07:00
Jim Fehlig
84a6209d7f libxl: queue shutdown event on domain shutdown
Emit libvirt shutdown event when receiving LIBXL_SHUTDOWN_REASON_POWEROFF
event from libxl.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-20 15:50:06 -07:00
John Ferlan
7eb37a0d65 bandwidth: Adjust documentation
Recent autotest/virt-test testing on f20 discovered an anomaly in how
the bandwidth options are documented and used. This was discovered due
to a bug fix in the /sbin/tc utility found in iproute-3.11.0.1 (on f20)
in which overflow was actually caught and returned as an error. The fix
was first introduced in iproute-3.10 (search on iproute2 commit 'a303853e').

The autotest/virt-test test for virsh domiftune was attempting to send
the largest unsigned integer value (4294967295) for maximum value
testing. The libvirt xml implementation was designed to manage values
in kilobytes thus when this value was passed to /sbin/tc, it (now)
properly rejected the 4294967295kbps value.

Investigation of the problem discovered that formatdomain.html.in and
formatnetwork.html.in described the elements and property types slightly
differently, although they use the same code - virNetDevBandwidthParseRate()
(shared by portgroups, domains, and networks xml parsers). Rather than
have the descriptions in two places, this patch will combine and reword
the description under formatnetwork.html.in and have formatdomain.html.in
link to that description.

This documentation faux pas was continued into the virsh man page where
the bandwidth description for both 'attach-interface' and 'domiftune'
did not indicate the format of each value, thus leading to the test using
largest unsigned integer value assuming "bps" rather than "kbps", which
ultimately was wrong.
2014-02-20 14:53:36 -05:00
Jim Fehlig
d716d942e2 libxl: always use libxlVmCleanupJob in shutdown thread
Commit e4a0e900 missed calling libxlVmCleanupJob in the shutdown
handler when processing a reboot event.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-20 11:50:33 -07:00
Eric Blake
60f7303c15 qemu: adjust maxmem/maxvcpu computation
https://bugzilla.redhat.com/show_bug.cgi?id=1038363

If a domain has a different maximum for persistent and live maxmem
or max vcpus, then it is possible to hit cases where libvirt
refuses to adjust the current values or gets halfway through
the adjustment before failing.  Better is to determine up front
if the change is possible for all requested flags.

Based on an idea by Geoff Franks.

* src/qemu/qemu_driver.c (qemuDomainSetMemoryFlags): Compute
correct maximum if both live and config are being set.
(qemuDomainSetVcpusFlags): Likewise.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 11:27:16 -07:00
Roman Bogorodskiy
2d2f29d232 Add Roman Bogorodskiy to the committers list 2014-02-20 20:51:08 +04:00
Daniel P. Berrange
432a3fee3b Rename virDomainGetRootFilesystem to virDomainGetFilesystemForTarget
The virDomainGetRootFilesystem method can be generalized to allow
any filesystem path to be obtained.

While doing this, start a new test case for purpose of testing various
helper methods in the domain_conf.{c,h} files, such as this one.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-20 15:50:46 +00:00
Daniel P. Berrange
590029f672 Introduce new OOM testing support
The previous OOM testing support would re-run the entire "main"
method each iteration, failing a different malloc each time.
When a test suite has 'n' allocations, the number of repeats
requires is  (n * (n + 1) ) / 2.  This gets very large, very
quickly.

This new OOM testing support instead integrates at the
virtTestRun level, so each individual test case gets repeated,
instead of the entire test suite. This means the values of
'n' are orders of magnitude smaller.

The simple usage is

   $ VIR_TEST_OOM=1 ./qemuxml2argvtest
   ...
   29) QEMU XML-2-ARGV clock-utc                                         ... OK
       Test OOM for nalloc=36 .................................... OK
   30) QEMU XML-2-ARGV clock-localtime                                   ... OK
       Test OOM for nalloc=36 .................................... OK
   31) QEMU XML-2-ARGV clock-france                                      ... OK
       Test OOM for nalloc=38 ...................................... OK
   ...

the second lines reports how many mallocs have to be failed, and thus
how many repeats of the test will be run.

If it crashes, then running under valgrind will often show the problem

  $ VIR_TEST_OOM=1 ../run valgrind ./qemuxml2argvtest

When debugging problems it is also helpful to select an individual
test case

  $ VIR_TEST_RANGE=30 VIR_TEST_OOM=1 ../run valgrind ./qemuxml2argvtest

When things get really tricky, it is possible to request that just
specific allocs are failed. eg to fail allocs 5 -> 12, use

  $ VIR_TEST_RANGE=30 VIR_TEST_OOM=1:5-12 ../run valgrind ./qemuxml2argvtest

In the worse case, you might want to know the stack trace of the
alloc which was failed then VIR_TEST_OOM_TRACE can be set. If it
is set to 1 then it will only print if it thinks a mistake happened.
This is often not reliable, so setting it to 2 will make it print
the stack trace for every alloc that is failed.

  $ VIR_TEST_OOM_TRACE=2 VIR_TEST_RANGE=30 VIR_TEST_OOM=1:5-5 ../run valgrind ./qemuxml2argvtest
  30) QEMU XML-2-ARGV clock-localtime                                   ... OK
      Test OOM for nalloc=36 !virAllocN
  /home/berrange/src/virt/libvirt/src/util/viralloc.c:180
  virHashCreateFull
  /home/berrange/src/virt/libvirt/src/util/virhash.c:144
  virDomainDefParseXML
  /home/berrange/src/virt/libvirt/src/conf/domain_conf.c:11745
  virDomainDefParseNode
  /home/berrange/src/virt/libvirt/src/conf/domain_conf.c:12646
  virDomainDefParse
  /home/berrange/src/virt/libvirt/src/conf/domain_conf.c:12590
  testCompareXMLToArgvFiles
  /home/berrange/src/virt/libvirt/tests/qemuxml2argvtest.c:106
  virtTestRun
  /home/berrange/src/virt/libvirt/tests/testutils.c:250
  mymain
  /home/berrange/src/virt/libvirt/tests/qemuxml2argvtest.c:418 (discriminator 2)
  virtTestMain
  /home/berrange/src/virt/libvirt/tests/testutils.c:750
  ??
  ??:0
  _start
  ??:?
   FAILED

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-20 15:36:10 +00:00
Daniel P. Berrange
cb9b3bc257 Fix multiple bugs in LXC domainMemoryStats driver
The virCgroupXXX APIs' return value must be checked for
being less than 0, not equal to 0.

An VIR_ERR_OPERATION_INVALID error must also be raised
when the VM is not running to prevent a crash on NULL
priv->cgroup field.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-02-20 15:32:49 +00:00
Thorsten Behrens
5eb61e6846 Add unit test for virCgroupGetPercpuStats. 2014-02-20 16:20:10 +01:00
Thorsten Behrens
35aebf1118 Fix misspelled cpuacct.usage_percpu in cgroup mock. 2014-02-20 16:20:10 +01:00
Thorsten Behrens
977a187ad7 Add unit test for virCgroupGetMemoryUsage. 2014-02-20 16:20:10 +01:00
Thorsten Behrens
9ce8a7c5c5 Add unit test for virCgroupGetBlkioIo*Serviced 2014-02-20 16:20:09 +01:00
Thorsten Behrens
0bd2ccdecc Widening API change - accept empty path for virDomainBlockStats
And provide domain summary stat in that case, for lxc backend.
Use case is a container inheriting all devices from the host,
e.g. when doing application containerization.
2014-02-20 16:20:09 +01:00
Thorsten Behrens
dcc85c603e Implement lxcDomainBlockStats* for lxc driver
Adds lxcDomainBlockStatsFlags and lxcDomainBlockStats functions.
2014-02-20 16:20:09 +01:00
Thorsten Behrens
4b3b2f6ceb Implement domainGetCPUStats for lxc driver. 2014-02-20 16:20:09 +01:00
Thorsten Behrens
65158899b7 Make qemuGetDomainTotalCPUStats a virCgroup function.
To reuse this from other drivers, like lxc.
2014-02-20 16:20:09 +01:00
Thorsten Behrens
192604ddee Implement domainMemoryStats API slot for LXC driver. 2014-02-20 16:20:09 +01:00
Thorsten Behrens
a2bb187c7e Add util virCgroupGetBlkioIo*Serviced methods.
This reads blkio stats from blkio.throttle.io_service_bytes and
blkio.throttle.io_serviced.
2014-02-20 16:20:09 +01:00
Jincheng Miao
6c1059ef24 virsh: fix memleak when starting a guest with invalid fd
When start a guest with --pass-fd, if the argument of --pass-fd is invalid,
virsh will exit, but doesn't free the variable 'dom'.

The valgrind said:
...
==24569== 63 (56 direct, 7 indirect) bytes in 1 blocks are definitely lost in loss record 130 of 234
==24569==    at 0x4C2A1D4: calloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==24569==    by 0x4E879A4: virAllocVar (viralloc.c:544)
==24569==    by 0x4EBD625: virObjectNew (virobject.c:190)
==24569==    by 0x4F3A18A: virGetDomain (datatypes.c:226)
==24569==    by 0x4F9311F: remoteDomainLookupByName (remote_driver.c:6636)
==24569==    by 0x4F44F20: virDomainLookupByName (libvirt.c:2277)
==24569==    by 0x12F616: vshCommandOptDomainBy (virsh-domain.c:105)
==24569==    by 0x131C79: cmdStart (virsh-domain.c:3330)
==24569==    by 0x12C4AB: vshCommandRun (virsh.c:1752)
==24569==    by 0x127001: main (virsh.c:3218)

https://bugzilla.redhat.com/show_bug.cgi?id=1067338

Signed-off-by: Jincheng Miao <jmiao@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
2014-02-20 05:40:13 -07:00
Richard Weinberger
39aad72510 lxc: Add destroy support for suspended domains
Destroying a suspended domain needs special action.
We cannot simply terminate all process because they are frozen.
Do deal with that we send them SIGKILL and thaw them.
Upon wakeup the process sees the pending signal and dies immediately.

Signed-off-by: Richard Weinberger <richard@nod.at>
2014-02-20 10:46:31 +01:00
Ján Tomko
057d26b2ac Fix build of portallocator on mingw
IN6ADDR_ANY_INIT does not seem to be working as expected on MinGW:
error: missing braces around initializer [-Werror=missing-braces]
         .sin6_addr = IN6ADDR_ANY_INIT,

Use the in6addr_any variable instead.

Reported by Daniel P. Berrange.
2014-02-20 10:16:07 +01:00
Michal Privoznik
83c404ff9b networkRunHook: Run hook only if possible
Currently, networkRunHook() is called in networkAllocateActualDevice and
friends. These functions, however, doesn't necessarily work on networks,
For example, if domain's interface is defined in this fashion:

    <interface type='bridge'>
      <mac address='52:54:00:0b:3b:16'/>
      <source bridge='virbr1'/>
      <model type='rtl8139'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x09' function='0x0'/>
    </interface>

The networkAllocateActualDevice jumps directly onto 'validate' label as
the interface is not type of 'network'. Hence, @network is left
initialized to NULL and networkRunHook(network, ...) is called. One of
the things that the hook function does is dereference @network. Soupir.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-02-20 08:56:17 +01:00
Jim Fehlig
e6dcb0e2a1 libxl: use job functions in libxlDomainSetSchedulerParametersFlags
Modify operation that needs to wait in the queue of modify jobs.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:01 -07:00
Jim Fehlig
7d9ff81603 libxl: use job functions in libxlDomainSetAutostart
Setting autostart is a modify operation that needs to wait in the
queue of modify jobs.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:01 -07:00
Jim Fehlig
85ff3d7aec libxl: use job functions in device attach and detach functions
These operations aren't necessarily time consuming, but need to
wait in the queue of modify jobs.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:01 -07:00
Jim Fehlig
7df46cff6b libxl: use job functions in vcpu set and pin functions
These operations aren't necessarily time consuming, but need to
wait in the queue of modify jobs.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:01 -07:00
Jim Fehlig
f9e6b7024c libxl: use job functions in libxlDomainCoreDump
Dumping a domain's core can take considerable time.  Use the
recently added job functions and unlock the virDomainObj while
dumping core.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:00 -07:00
Jim Fehlig
341870b10d libxl: use job functions in domain save operations
Saving domain memory and cpu state can take considerable time.
Use the recently added job functions and unlock the virDomainObj
while saving the domain.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:00 -07:00
Jim Fehlig
e4a0e900d3 libxl: use job functions when cleaning up a domain
When explicitly destroying a domain (libxlDomainDestroyFlags), or
handling an out-of-band domain shutdown event, cleanup the domain
in the context of a job.  Introduce libxlVmCleanupJob to wrap
libxlVmCleanup in a job block.
2014-02-19 11:10:00 -07:00
Jim Fehlig
f5bc5bd4df libxl: use job functions in libxlDomain{Suspend,Resume}
These operations aren't necessarily time consuming, but need to
wait in the queue of modify jobs.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:00 -07:00
Jim Fehlig
ac1444c35f libxl: use job functions in libxlDomainSetMemoryFlags
Large balloon operation can be time consuming.  Use the recently
added job functions and unlock the virDomainObj while ballooning.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
2014-02-19 11:10:00 -07:00