KubeVirt uses libvirt to start qrap in its current draft integration (https://github.com/kubevirt/kubevirt/pull/7849/), and libvirtd starts qrap three times every time a new virtual machine is created: once on domain creation, and twice on domain start (for "probing") and to finally start it for real. Very often, a subsequent invocation of qrap happen before the previously running instance of qemu terminates, which means that passt will refuse the new connection as the old one is still active. Introduce a single retry with a 100ms delay to work around this. This should be checked again once native libvirt support is there, and that point qrap will have no reason to exist anymore. Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com> --- qrap.c | 34 +++++++++++++++++++++++++++++++++- 1 file changed, 33 insertions(+), 1 deletion(-) diff --git a/qrap.c b/qrap.c index 17cc472..303e981 100644 --- a/qrap.c +++ b/qrap.c @@ -25,6 +25,7 @@ #include <net/if_arp.h> #include <netinet/in.h> #include <netinet/if_ether.h> +#include <time.h> #include "util.h" #include "passt.h" @@ -112,7 +113,7 @@ void usage(const char *name) int main(int argc, char **argv) { struct timeval tv = { .tv_sec = 0, .tv_usec = (long)(500 * 1000) }; - int i, s, qemu_argc = 0, addr_map = 0, has_dev = 0; + int i, s, qemu_argc = 0, addr_map = 0, has_dev = 0, retry_on_reset; char *qemu_argv[ARG_MAX], dev_str[ARG_MAX]; struct sockaddr_un addr = { .sun_family = AF_UNIX, @@ -233,6 +234,9 @@ int main(int argc, char **argv) valid_args: for (i = 1; i < UNIX_SOCK_MAX; i++) { + retry_on_reset = 1; + +retry: s = socket(AF_UNIX, SOCK_STREAM, 0); if (s < 0) { perror("socket"); @@ -254,6 +258,34 @@ valid_args: else break; + /* FIXME: in a KubeVirt environment, libvirtd invokes qrap three + * times in a strict sequence when a virtual machine needs to + * be started, namely, when: + * - the domain XML is saved + * - the domain is started (for "probing") + * - the virtual machine is started for real + * and it often happens that the qemu process is still running + * when qrap is invoked again, so passt will refuse the new + * connection because the previous one is still active. This + * overlap seems to be anywhere between 0 and 3ms. + * + * If we get a connection reset, retry, just once, after 100ms, + * to allow for the previous qemu instance to terminate and, in + * turn, for the connection to passt to be closed. + * + * This should be fixed in libvirt instead. It probably makes + * sense to check this behaviour once native libvirt support is + * there, and this implies native qemu support too, so at that + * point qrap will have no reason to exist anymore -- that is, + * this FIXME will probably remain until the tool itself is + * obsoleted. + */ + if (retry_on_reset && errno == ECONNRESET) { + retry_on_reset = 0; + usleep(100 * 1000); + goto retry; + } + fprintf(stderr, "Probe of %s failed\n", addr.sun_path); close(s); -- 2.35.1