- Devel - libvirt List Archives

[libvirt] libvirtd can not start up
by Lin Qing 16 Jun '12

16 Jun '12

Hi all: I met a issue during boot libvirtd.When I use "libvirtd --daemon " , libvirtd didn't start up. And I am using the latest libvirt source code . Here is the libvirtd.log /usr/local/var/log/libvirt/libvirtd.log 2012-06-15 06:35:04.230+0000: 16193: info : libvirt version: 0.9.12 2012-06-15 06:35:04.230+0000: 16193: error : virStateInitialize:781 : Initialization of NWFilter state driver failed 2012-06-15 06:35:05.079+0000: 16193: error : daemonRunStateInit:752 : Driver state initialization failed anybody konws what 's going on ? Thanks Lin Qing.

3 4

[libvirt] libvirtd stopped launching
by Zeeshan Ali (Khattak) 16 Jun '12

16 Jun '12

Hi everyone, After I updated my libvirt repo/install from latest git master yesterday, libvirtd won't just launch for me anymore. The error in log was: 2012-06-13 11:59:47.427+0000: 30501: error : virStateInitialize:781 : Initialization of NWFilter state driver failed 2012-06-13 11:59:48.601+0000: 30501: error : daemonRunStateInit:752 : Driver state initialization failed Daniel asked me to enable more debug to find the cause and here is the output: http://static.fi/~zeenix/tmp/libvirt.out Does anyone know whats going on and how can I fix this? In the meantime I'll try to see reverting to some older commit helps and if so, I'll do a git bisect.. -- Regards, Zeeshan Ali (Khattak) FSF member#5124

3 3

[libvirt] Website sidebar not consistent between libvirt.org and wiki.libvirt.org
by Cole Robinson 15 Jun '12

15 Jun '12

The sidebar here looks right: http://libvirt.org/ But the sidebar here seems pretty out of date: http://wiki.libvirt.org/ Including missing options, and the FAQ link pointing to the old FAQ.html file which shouldn't exist anymore. Any one know how to fix? Thanks, Cole

2 4

[libvirt] [PATCH] docs: Update domain XML schema
by Osier Yang 15 Jun '12

15 Jun '12

'boot' tag shouldn't be exclusive with 'kernel', 'initrd', and 'cmdline', though the boot sequence doesn't make sense when the guest boots from kernel directly. But it's useful if booting from kernel is to install a newly guest, even it's not to install a guest, there is no hurt. And on the other hand, we allow 'boot' and the kernel tags when parsing. --- docs/schemas/domaincommon.rng | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 62c28c8..46e539d 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -222,12 +222,14 @@ <ref name="absFilePath"/> </element> </optional> - <choice> + <optional> <ref name="osbootkernel"/> + </optional> + <optional> <zeroOrMore> <ref name="osbootdev"/> </zeroOrMore> - </choice> + </optional> <optional> <element name="bootmenu"> <attribute name="enable"> -- 1.7.7.3

2 2

[libvirt] [PATCH] Add more debug logging for libvirtd startup
by Daniel P. Berrange 15 Jun '12

15 Jun '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> To facilitate future troubleshooting add a bunch more debugging statements into important startup parts of libvirt --- daemon/libvirtd.c | 40 +++++++++++++++++++++++++++++----------- src/libvirt.c | 15 +++++++++------ src/util/hooks.c | 3 +++ src/util/virnetlink.c | 1 + 4 files changed, 42 insertions(+), 17 deletions(-) diff --git a/daemon/libvirtd.c b/daemon/libvirtd.c index c1ee3f4..9c06344 100644 --- a/daemon/libvirtd.c +++ b/daemon/libvirtd.c @@ -461,6 +461,7 @@ static int daemonSetupNetworking(virNetServerPtr srv, goto error; } + VIR_DEBUG("Registering unix socket %s", sock_path); if (!(svc = virNetServerServiceNewUNIX(sock_path, unix_sock_rw_mask, unix_sock_gid, @@ -469,15 +470,17 @@ static int daemonSetupNetworking(virNetServerPtr srv, config->max_client_requests, NULL))) goto error; - if (sock_path_ro && - !(svcRO = virNetServerServiceNewUNIX(sock_path_ro, - unix_sock_ro_mask, - unix_sock_gid, - config->auth_unix_ro, - true, - config->max_client_requests, - NULL))) - goto error; + if (sock_path_ro) { + VIR_DEBUG("Registering unix socket %s", sock_path_ro); + if (!(svcRO = virNetServerServiceNewUNIX(sock_path_ro, + unix_sock_ro_mask, + unix_sock_gid, + config->auth_unix_ro, + true, + config->max_client_requests, + NULL))) + goto error; + } if (virNetServerAddService(srv, svc, config->mdns_adv && !ipsock ? @@ -491,6 +494,8 @@ static int daemonSetupNetworking(virNetServerPtr srv, if (ipsock) { if (config->listen_tcp) { + VIR_DEBUG("Registering TCP socket %s:%s", + config->listen_addr, config->tcp_port); if (!(svcTCP = virNetServerServiceNewTCP(config->listen_addr, config->tcp_port, config->auth_tcp, @@ -527,6 +532,8 @@ static int daemonSetupNetworking(virNetServerPtr srv, goto error; } + VIR_DEBUG("Registering TLS socket %s:%s", + config->listen_addr, config->tls_port); if (!(svcTLS = virNetServerServiceNewTCP(config->listen_addr, config->tls_port, @@ -1070,8 +1077,10 @@ int main(int argc, char **argv) { } if (!privileged && - migrateProfile() < 0) + migrateProfile() < 0) { + VIR_ERROR(_("Exiting due to failure to migrate profile")); exit(EXIT_FAILURE); + } if (config->host_uuid && virSetHostUUIDStr(config->host_uuid) < 0) { @@ -1090,6 +1099,7 @@ int main(int argc, char **argv) { VIR_ERROR(_("Can't determine pid file path.")); exit(EXIT_FAILURE); } + VIR_DEBUG("Decided on pid file path '%s'", NULLSTR(pid_file)); if (daemonUnixSocketPaths(config, privileged, @@ -1098,6 +1108,8 @@ int main(int argc, char **argv) { VIR_ERROR(_("Can't determine socket paths")); exit(EXIT_FAILURE); } + VIR_DEBUG("Decided on socket paths '%s' and '%s'", + sock_file, NULLSTR(sock_file_ro)); if (godaemon) { char ebuf[1024]; @@ -1135,6 +1147,7 @@ int main(int argc, char **argv) { old_umask = umask(022); else old_umask = umask(077); + VIR_DEBUG("Ensuring run dir '%s' exists", run_dir); if (virFileMakePath(run_dir) < 0) { char ebuf[1024]; VIR_ERROR(_("unable to create rundir %s: %s"), run_dir, @@ -1171,6 +1184,7 @@ int main(int argc, char **argv) { /* Beyond this point, nothing should rely on using * getuid/geteuid() == 0, for privilege level checks. */ + VIR_DEBUG("Dropping privileges (if required)"); if (daemonSetupPrivs() < 0) { ret = VIR_DAEMON_ERR_PRIVS; goto cleanup; @@ -1207,11 +1221,13 @@ int main(int argc, char **argv) { goto cleanup; } - if (timeout != -1) + if (timeout != -1) { + VIR_DEBUG("Registering shutdown timeout %d", timeout); virNetServerAutoShutdown(srv, timeout, daemonShutdownCheck, NULL); + } if ((daemonSetupSignals(srv)) < 0) { ret = VIR_DAEMON_ERR_SIGNAL; @@ -1219,11 +1235,13 @@ int main(int argc, char **argv) { } if (config->audit_level) { + VIR_DEBUG("Attempting to configure auditing subsystem"); if (virAuditOpen() < 0) { if (config->audit_level > 1) { ret = VIR_DAEMON_ERR_AUDIT; goto cleanup; } + VIR_DEBUG("Proceeding without auditing"); } } virAuditLog(config->audit_logging); diff --git a/src/libvirt.c b/src/libvirt.c index 16afd58..99e85e4 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -769,20 +769,23 @@ virRegisterStateDriver(virStateDriverPtr driver) * Returns 0 if all succeed, -1 upon any failure. */ int virStateInitialize(int privileged) { - int i, ret = 0; + int i; if (virInitialize() < 0) return -1; for (i = 0 ; i < virStateDriverTabCount ; i++) { - if (virStateDriverTab[i]->initialize && - virStateDriverTab[i]->initialize(privileged) < 0) { - VIR_ERROR(_("Initialization of %s state driver failed"), + if (virStateDriverTab[i]->initialize) { + VIR_DEBUG("Running global init for %s state driver", virStateDriverTab[i]->name); - ret = -1; + if (virStateDriverTab[i]->initialize(privileged) < 0) { + VIR_ERROR(_("Initialization of %s state driver failed"), + virStateDriverTab[i]->name); + return -1; + } } } - return ret; + return 0; } /** diff --git a/src/util/hooks.c b/src/util/hooks.c index f89a40f..e0d6bf0 100644 --- a/src/util/hooks.c +++ b/src/util/hooks.c @@ -271,6 +271,9 @@ virHookCall(int driver, return -1; } + VIR_DEBUG("Calling hook opstr=%s subopstr=%s extra=%s", + opstr, subopstr, extra); + cmd = virCommandNewArgList(path, id, opstr, subopstr, extra, NULL); virCommandAddEnvPassCommon(cmd); diff --git a/src/util/virnetlink.c b/src/util/virnetlink.c index 9539694..bb0dae9 100644 --- a/src/util/virnetlink.c +++ b/src/util/virnetlink.c @@ -130,6 +130,7 @@ virNetlinkStartup(void) { if (placeholder_nlhandle) return 0; + VIR_DEBUG("Running global netlink initialization"); placeholder_nlhandle = virNetlinkAlloc(); if (!placeholder_nlhandle) { virReportSystemError(errno, "%s", -- 1.7.10.2

2 1

[libvirt] [PATCH] Fix launch of libvirtd when DHCP snooping support is not available.
by Daniel P. Berrange 15 Jun '12

15 Jun '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> When libpcap is not available, the NWFilter driver provides a no-op stub for the DHCP snooping initialization. This was mistakenly returning '-1' instead of '0', so the entire driver initialization failed --- src/nwfilter/nwfilter_dhcpsnoop.c | 5 ++++- src/nwfilter/nwfilter_gentech_driver.c | 1 + src/nwfilter/nwfilter_learnipaddr.c | 1 + 3 files changed, 6 insertions(+), 1 deletion(-) diff --git a/src/nwfilter/nwfilter_dhcpsnoop.c b/src/nwfilter/nwfilter_dhcpsnoop.c index 22f76e2..5dffaad 100644 --- a/src/nwfilter/nwfilter_dhcpsnoop.c +++ b/src/nwfilter/nwfilter_dhcpsnoop.c @@ -2056,6 +2056,8 @@ virNWFilterDHCPSnoopInit(void) if (virNWFilterSnoopState.snoopReqs) return 0; + VIR_DEBUG("Initializing DHCP snooping"); + if (virMutexInitRecursive(&virNWFilterSnoopState.snoopLock) < 0 || virMutexInit(&virNWFilterSnoopState.activeLock) < 0 || virAtomicIntInit(&virNWFilterSnoopState.nLeases) < 0 || @@ -2176,7 +2178,8 @@ virNWFilterDHCPSnoopShutdown(void) int virNWFilterDHCPSnoopInit(void) { - return -1; + VIR_DEBUG("No DHCP snooping support available"); + return 0; } void diff --git a/src/nwfilter/nwfilter_gentech_driver.c b/src/nwfilter/nwfilter_gentech_driver.c index 4769d21..d833bbb 100644 --- a/src/nwfilter/nwfilter_gentech_driver.c +++ b/src/nwfilter/nwfilter_gentech_driver.c @@ -57,6 +57,7 @@ static virNWFilterTechDriverPtr filter_tech_drivers[] = { void virNWFilterTechDriversInit(bool privileged) { int i = 0; + VIR_DEBUG("Initializing NWFilter technology drivers"); while (filter_tech_drivers[i]) { if (!(filter_tech_drivers[i]->flags & TECHDRV_FLAG_INITIALIZED)) filter_tech_drivers[i]->init(privileged); diff --git a/src/nwfilter/nwfilter_learnipaddr.c b/src/nwfilter/nwfilter_learnipaddr.c index af13738..a891377 100644 --- a/src/nwfilter/nwfilter_learnipaddr.c +++ b/src/nwfilter/nwfilter_learnipaddr.c @@ -782,6 +782,7 @@ virNWFilterLearnInit(void) { if (pendingLearnReq) return 0; + VIR_DEBUG("Initializing IP address learning"); threadsTerminate = false; pendingLearnReq = virHashCreate(0, freeLearnReqEntry); -- 1.7.10.2

2 1

[libvirt] RFC: spice-server default listen behaviour change
by David Jaša 15 Jun '12

15 Jun '12

Hi, I learned few things about ipv6 lately. Most importantly about dual-socket that means that a process that opens ::0 automatically listens on both ipv4 _and_ ipv6 unless it sets IPV6_ONLY option of setsockopt() to 0. This is pretty important wrt dual-stack configurations because they can be implemented with just slight changes to spice server (unlike the old RFE requesting listening on multiple addresses): * when no addr= or ipvx options are set, listen on ::0 * when ipv4 and no addr= option is set, listen on 0.0.0.0 * when ipv6 is set, set IPV6_ONLY to 1 to make sure that spice server won't listen on ipv4 * when conflicting ipvx and addr= options are set, error out (this already works fine) This will affect upper layers though, given bugs like https://bugzilla.redhat.com/show_bug.cgi?id=832121 , but it seems to me like the step in the right direction. Any thoughts/comments before I file this as a bug? David -- David Jaša, RHCE SPICE QE based in Brno GPG Key: 22C33E24 Fingerprint: 513A 060B D1B4 2A72 7F0D 0278 B125 CD00 22C3 3E24

3 11

[libvirt] [PATCH 0/3 v5] use qemu's dump-guest-meory when vm uses host device
by Wen Congyang 15 Jun '12

15 Jun '12

Currently, we use migrate to dump guest's memory. There is one restriction in migrate command: the device's status should be stored in qemu because the device's status should be passed to target machine. If we passthrough a host device to guest, the device's status is stored in the real device. So migrate command will fail. We usually use dump when guest is panicked. So there is no need to store device's status in the vmcore. qemu have a new monitor command dump-guest-memory to dump guest memory, but it doesn't support async now(it will support later when the common async API is implemented). Changes from v4 to v5: 1. address Martin Kletzander's comment Changes from v3 to v4: 1. allow the user to specify the core file's format Changes from v2 to v3: 1. qemu supports the fd that is associated with a pipe, socket, or FIFO. So pass a pipe fd to qemu and O_DIRECT can work now. Change from v1 to v2: 1. remove the implemention for text mode. Wen Congyang (3): qemu: implement qemu's dump-guest-memory qemu: allow the client to choose the vmcore's format virsh: allow the user to specify vmcore's format include/libvirt/libvirt.h.in | 1 + src/qemu/qemu_capabilities.c | 5 +++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_domain.c | 1 + src/qemu/qemu_domain.h | 1 + src/qemu/qemu_driver.c | 60 ++++++++++++++++++++++++++++++++--------- src/qemu/qemu_monitor.c | 37 ++++++++++++++++++++++++++ src/qemu/qemu_monitor.h | 12 ++++++++ src/qemu/qemu_monitor_json.c | 34 +++++++++++++++++++++++ src/qemu/qemu_monitor_json.h | 6 ++++ tools/virsh.c | 3 ++ tools/virsh.pod | 5 +++- 12 files changed, 152 insertions(+), 14 deletions(-)

3 8

[libvirt] [PATCH] qemu: fix potential dead lock
by Wen Congyang 15 Jun '12

15 Jun '12

If we lock the qemu_driver, we should call qemuDomainObjBeginJobWithDriver() not qemuDomainObjBeginJob(). --- src/qemu/qemu_driver.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 39b27b1..3d62aab 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -8905,7 +8905,7 @@ qemuDomainMigrateBegin3(virDomainPtr domain, goto cleanup; asyncJob = QEMU_ASYNC_JOB_MIGRATION_OUT; } else { - if (qemuDomainObjBeginJob(driver, vm, QEMU_JOB_MODIFY) < 0) + if (qemuDomainObjBeginJobWithDriver(driver, vm, QEMU_JOB_MODIFY) < 0) goto cleanup; asyncJob = QEMU_ASYNC_JOB_NONE; } -- 1.7.1

3 2

[libvirt] RPC ref count question
by Radu Caragea 15 Jun '12

15 Jun '12

Hi everyone, >From what I understand after defining a domain with remoteDomainCreateXML, the corresponding destroy call is remoteDomainDestroy, both from the automatically generated remote_client_bodies.h My question is why the domain isn't freed in the destroy procedure (with virUnrefDomain or other) ? Is it supposed to be called separately or is it a bug ? I'm asking because from what I saw the connection refcount increments in domain creation but doesn't decrement in destroy.

2 1