October 2012 - Devel - libvirt List Archives

[libvirt] IPV6 and routing
by Gene Czarcinski 09 Oct '12

09 Oct '12

As I have mentioned in other messages, I am interested in having full support for IPv6 in libvirt. To me this includes having dhcp6 for IPV6 address assignment and using RA (radvd) to establish the default route. This is what I am using on my real LANs. Before starting into adding dhcp6 support to libvirt, I wanted to see just how it works with the current software. First of all, it appears that, when nat or routed are specified for IPv4, the IPv6 is routed. If it is an isolated/private network, then it can only work with other guests on that network. The iptables and ip6tables settings corresponded and were as expected. On the virtualization host, both IPv4 and IPv6 forwarding are enabled. While I can easily do stuff like ping6 and ssh -6 from virtual guests to the virtualization host, I have been unable to do anything with external hosts ... unless I add a static route for the virtual IPv6 network on the target host back to the virtualization host. This is the only way I have gotten anything to work. To complicate things, it seem that "everything" wants the IPv6 network to have prefix=64 or things do not work correctly. The real systems use fd00:dead:beef:17::/64 for their network. The virtual networks all use fd00:face:17:xx::/64 for their networks. The network traffic on the virtualization host is forwarded to the target host ... I can see the packets with wireshark on the target host. On the target host I tried specifying a static route for network fd00:face:17::/48 ... well, that really screwed things up, resulted in some "redirects" from the virtualization host saying the that it was sent a malformed packed ... it took a reboot to clean things up. OK, so leave the fd00:face:17:6::/64 static route on the target host but subnetwork this network on the virtualization host using networks like fd00:face:17:6:8::/80 and fd00:face:6:9::/80. This works if I manually configure IPv6 on the virtual guest. Since radvd is "upset" by a non-prefix=64 network, I was not surprised when the guest's automatic IPv6 address/network was not configured. OK, what am I missing? What don't I understand? If IPv6 is going to be useful in virtualization, then there must be some "easy" way to have other systems understand that the virtualization host is acting as a router for the virtual IPv6 networks it runs. While being able to go between the virtualization hosts and the virtual guests is very useful, I do not consider this sufficient. I have googled and found some stuff as well as reading more RFCs than I wanted to but I cannot find anything to address this issue. IIRC, I did find something under a libvirt document that indicates "routed" will be used for some kind of subnetworking. Does libvirt need an IPv6 "NAT" to make this work? Comments? Suggestions? Gene

3 5

[libvirt] ANNOUNCE: libvirt 0.9.11.6 maintenance release
by Cole Robinson 09 Oct '12

09 Oct '12

libvirt 0.9.11.6 maintenance release is now available. This is libvirt 0.9.11 with additional bugfixes that have accumulated upstream since the initial release. This release can be downloaded at: http://libvirt.org/sources/stable_updates/libvirt-0.9.11.6.tar.gz Changes in this version: * remove dnsmasq command line parameter "--filterwin2k" * dnsmasq: avoid forwarding queries without a domain * security: Fix libvirtd crash possibility For info about past maintenance releases, see: http://wiki.libvirt.org/page/Maintenance_Releases Thanks, Cole

3 3

[libvirt] [PATCH] hooks: let virCommand do the error reporting
by Eric Blake 09 Oct '12

09 Oct '12

The code was reporting raw exit status without decoding it into normal vs. signal exit. virCommandRun already does this, but with a different error type, so all we have to do is recast the error to the correct type. Reported by li guang. * src/util/hooks.c (virHookCall): Simplify. --- In response to: https://www.redhat.com/archives/libvir-list/2012-October/msg00308.html src/util/hooks.c | 14 ++++++-------- 1 file changed, 6 insertions(+), 8 deletions(-) diff --git a/src/util/hooks.c b/src/util/hooks.c index f5890d2..8817a4e 100644 --- a/src/util/hooks.c +++ b/src/util/hooks.c @@ -1,7 +1,7 @@ /* * hooks.c: implementation of the synchronous hooks support * - * Copyright (C) 2010-2011 Red Hat, Inc. + * Copyright (C) 2010-2012 Red Hat, Inc. * Copyright (C) 2010 Daniel Veillard * * This library is free software; you can redistribute it and/or @@ -206,7 +206,6 @@ virHookCall(int driver, char **output) { int ret; - int exitstatus; char *path; virCommandPtr cmd; const char *drvstr; @@ -279,12 +278,11 @@ virHookCall(int driver, if (output) virCommandSetOutputBuffer(cmd, output); - ret = virCommandRun(cmd, &exitstatus); - if (ret == 0 && exitstatus != 0) { - virReportError(VIR_ERR_HOOK_SCRIPT_FAILED, - _("Hook script %s %s failed with error code %d"), - path, drvstr, exitstatus); - ret = -1; + ret = virCommandRun(cmd, NULL); + if (ret < 0) { + /* Convert INTERNAL_ERROR into known error. */ + virErrorPtr err = virGetLastError(); + virReportError(VIR_ERR_HOOK_SCRIPT_FAILED, "%s", err->message); } virCommandFree(cmd); -- 1.7.11.4

2 3

[libvirt] [libvirt-designer][PATCH 1/2] spec: Update changelog
by Michal Privoznik 09 Oct '12

09 Oct '12

to reflect the brand new release --- libvirt-designer.spec.in | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/libvirt-designer.spec.in b/libvirt-designer.spec.in index 667366c..2ff7487 100644 --- a/libvirt-designer.spec.in +++ b/libvirt-designer.spec.in @@ -103,3 +103,6 @@ rm -rf $RPM_BUILD_ROOT %{_datadir}/gtk-doc/html/libvirt-designer %changelog + +Tue Oct 09 2012 Daniel P. Berrange <berrange(a)redhat.com> 0.0.1-1 +- created -- 1.7.8.6

2 5

[libvirt] [PATCH] doc: update description about user/group in qemu.conf
by Marcelo Cerri 09 Oct '12

09 Oct '12

As a side effect of changes in the functions virGetUserID and virGetGroupID, the user and group configurations for DAC in qemu.conf are now able to accept both names and IDs, supporting a leading plus sign to ensure that a numeric value will not be interpreted as a name. This patch updates the comments in qemu.conf, including a description of this new behavior. --- src/qemu/qemu.conf | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index 18105ca..dd853c8 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -180,10 +180,25 @@ # guests will be blocked. Defaults to 0. #security_require_confined = 1 -# The user ID for QEMU processes run by the system instance. +# The user for QEMU processes run by the system instance. It can be +# specified as a user name or as a user id. The qemu driver will try to +# parse this value first as a name and then, if the name doesn't exist, +# as a user id. +# +# Since a sequence of digits is a valid user name, a leading plus sign +# can be used to ensure that a user id will not be interpreted as a user +# name. +# +# Some examples of valid values are: +# +# user = "qemu" # A user named "qemu" +# user = "+0" # Super user (uid=0) +# user = "100" # A user named "100" or a user with uid=100 +# #user = "root" -# The group ID for QEMU processes run by the system instance. +# The group for QEMU processes run by the system instance. It can be +# specified in a similar way to user. #group = "root" # Whether libvirt should dynamically change file ownership -- 1.7.12

2 1

[libvirt] [PATCH] nodeinfo: Fully convert to new virReportError
by Michal Privoznik 09 Oct '12

09 Oct '12

With our latest s/[a-z]+ReportError/virReportError/ rewrite (47ab34e2) we forgot to update arm part of the code. --- Thanks to my masochist idea of compiling libvirt on my Raspberry PI I've discovered this. Pushing under trivial & build breaker (on arm platforms obviously) rules. For those who are interested, it took ages and we don't build cleanly: http://pastebin.com/dkrJaV5Z But I am not sure how to fix those 'cast increases required alignment of target type' warnings. And I am not sure anybody else is ever going to use libvirt on arm. src/nodeinfo.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 4cbfe63..c0e60d8 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -423,8 +423,8 @@ int linuxNodeInfoCPUPopulate(FILE *cpuinfo, buf++; if (*buf != ':' || !buf[1]) { - nodeReportError(VIR_ERR_INTERNAL_ERROR, - "%s", _("parsing cpu MHz from cpuinfo")); + virReportError(VIR_ERR_INTERNAL_ERROR, + "%s", _("parsing cpu MHz from cpuinfo")); goto cleanup; } -- 1.7.8.6

3 2

[libvirt] [PATCH] docs: Drop useless </p> in drvphyp.html.in
by Michal Privoznik 09 Oct '12

09 Oct '12

--- Pushing under trivial rule. docs/drvphyp.html.in | 1 - 1 files changed, 0 insertions(+), 1 deletions(-) diff --git a/docs/drvphyp.html.in b/docs/drvphyp.html.in index 0bd20dd..50e26b2 100644 --- a/docs/drvphyp.html.in +++ b/docs/drvphyp.html.in @@ -43,6 +43,5 @@ phyp://user@ivm/system (IVM connection) <pre> phyp://[username@]{hmc|ivm}/managed_system </pre> - </p> </body></html> -- 1.7.8.6

1 0

[libvirt] [PATCH v2 0/2] extend virGetUserID and virGetGroupID
by Marcelo Cerri 09 Oct '12

09 Oct '12

This patch series moves the logic for parsing users and groups in a similar way to coreutils' chown from security_dac.c to util.c, as suggested by Eric Blake. This change has two majors side effects: 1. Some error messages that were issued when security_dac.c tried to parse an ID as a name are no longer issued. 2. The keys `user` and `group` in qemu.conf can now be defined in the same way that in DAC security labels. Peter Krempa's fix for correctly handling errors returned by getpwnam_r and getgrnam_r is squashed into this patch series. Marcelo Cerri (2): util: extend virGetUserID and virGetGroupID to support names and IDs security: update user and group parsing in security_dac.c src/security/security_dac.c | 45 +++----------- src/util/util.c | 143 ++++++++++++++++++++++++++++++++------------ 2 files changed, 112 insertions(+), 76 deletions(-) -- 1.7.12

2 4

[libvirt] [PATCH 0/3] Cleanly save session VMs on logout/shutdown
by Alexander Larsson 09 Oct '12

09 Oct '12

Atm, when you exit the sessin (via logout or shutdown for instance) we just leave the qemu:///session handling libvirtd process around, including any running VMs. This is not particularly nice, as there is no way to know that the VMs are running, and you risk data loss if you shutdown while the VMs are running. This patch series adds a babysitter to the session that will track session lifetime and shutdown events and save running VMs as needed. Outstanding questions: Can this code be inside libvirtd itself? I.E. is it possible to do VM management calls like virDomainManagedSave() from inside libvirtd itself. If not, who will launch the babysitter? Alexander Larsson (3): Fix typo in HAVE_DBUS automake conditional virdbus: Add virDBusGetSessionBus helper Add new libvirt-babysitter tool configure.ac | 2 +- src/util/virdbus.c | 84 ++++++++++---- src/util/virdbus.h | 1 + tools/Makefile.am | 21 ++++ tools/libvirt-babysitter.c | 276 +++++++++++++++++++++++++++++++++++++++++++++ 5 files changed, 363 insertions(+), 21 deletions(-) create mode 100644 tools/libvirt-babysitter.c -- 1.7.12.1

4 9

[libvirt] [PATCH 00/12] Structured syslog (Lumberjack/CEE) support
by Miloslav Trmač 09 Oct '12

09 Oct '12

This patch set adds structured syslog support to libvirt. In short, the idea is to embed JSON in a syslog record. This format is used in a new log output type called "syslog/json". Example "syslog/json" record (line-wrapped for readability): Sep 20 19:59:40 kulicka libvirt: @cee: {"msg":"libvirt version: 0.10.1", "category":"../../src/test/test_driver.c","priority":"info", "funcname":"virLogVMessage","line":769, "timestamp":"2012-09-20 17:59:40.853+0000"} compared with existing "syslog" record format: Sep 20 20:02:24 kulicka libvirt: 5896: info : libvirt version: 0.10.1 This JSON embedding format is defined by the Lumberjack project led by the major Open Source syslog implementations (https://fedorahosted.org/lumberjack/) The format is also a base for a future MITRE CEE standard. Some of the fields are specified at https://fedorahosted.org/lumberjack/wiki/FieldList , but applications are free to use whatever fields make sense for them. The JSON structure makes easy to process logs by programs without writing code that understands the specific format used by each application, or using unreliable regexps (https://fedorahosted.org/ceelog/ is a companion project that allows processing such logs). Current systems will store and these records unmodified and present them to users unmodified, but tools (and probably /var/log/messages) will be able to show only the "msg" field; this allows us to add more detailed information without worsening the user experience (the example above already shows added information that is currently being thrown away when logging to syslog). The patch set, aside from building the required infrastructure, has two main components: * Add the "syslog/json" output type, and use it to record all information passed to virLogMessage. The example result is shown above. This primarily makes it easier to parse the message. * Allow callers of virLogMessage to add arbitrary additional JSON fields, and use it in virRaiseErrorFull and its callers. Because the additional fields for errors include virErrorDomain and virErrorNumber, this allows _programs_ to _automatically_ detect the event and react to it, perhaps alerting the user about a situation that requires immediate attention, or automatically suggesting a solution. This has not been possible until now because the current logs contain only a translated error message, so the log record doesn't contain anything stable that could be detected by a program. It is assumed that other parts of libvirt may add more event-specific JSON fields in the future, based on demand by applications that process the logs; I have, however, no immediate plans to add more. Miloslav Trmač (12): Drop unused return value of virLogOutputFunc Add JSON API for preallocated objects. Split yajl_gen allocation into a separate function. Split adding object properties from virJSONValueToStringOne Add a JSON generator API. Pass the "raw" log message to each virLogOutputFunc Split priority conversion from virLogOutputToSyslog Add "syslog/json" log output format. Add a JSON properties parameter to virLog{,V}Message Add a JSON properties parameter to virLogOutputFunc Add detailed error information to JSON syslog Add a test for the new error reporting mechanism. docs/logging.html.in | 3 + src/libvirt_private.syms | 4 + src/uml/uml_conf.c | 2 +- src/util/json.c | 232 +++++++++++++++++++++++++++++++++++++++++----- src/util/json.h | 26 ++++++ src/util/logging.c | 233 ++++++++++++++++++++++++++++++++++------------- src/util/logging.h | 35 +++---- src/util/viraudit.c | 4 +- src/util/virterror.c | 32 ++++++- tests/Makefile.am | 6 +- tests/errorjsontest.c | 211 ++++++++++++++++++++++++++++++++++++++++++ tests/testutils.c | 7 +- 12 files changed, 684 insertions(+), 111 deletions(-) create mode 100644 tests/errorjsontest.c -- 1.7.11.4

4 24