On Wed, Mar 06, 2024 at 02:28:34AM -0800, Andrea Bolognani wrote: > On Tue, Mar 05, 2024 at 04:08:20PM -0500, Chuck Lever wrote: > > Hello - > > > > The virsh(1) man page says: > > > > > --nvram and --keep-nvram specify accordingly to delete or keep nvram (/domain/os/nvram/) file. > > > > However on my Fedora 39 system with libvirt 9.7.0, using "virsh > > undefine --nvram" option appears to leave the NVRAM backing file: > > > > cel@boudin:~$ ls ~/.config/libvirt/qemu/nvram > > guestfs-02a4oblo9vg7888d_VARS.qcow2 guestfs-7hmiolqcqwz5b0qs_VARS.qcow2 > > guestfs-el7lqymnwgzpnr38_VARS.qcow2 guestfs-jp0tfneh205xjrh8_VARS.qcow2 > > guestfs-mhqbp4gwmqsuu9w3_VARS.qcow2 guestfs-sat7x2kgpg9opjds_VARS.qcow2 > > guestfs-vhhi476rsq7vh83z_VARS.qcow2 guestfs-177mrq9xrb5589wl_VARS.qcow2 > > guestfs-88p8heyq1j0sy8bk_VARS.qcow2 guestfs-gxul3j9e1h3po19o_VARS.qcow2 > > guestfs-k0jf8msw4kh3yuff_VARS.qcow2 guestfs-oakcwpbozezvids8_VARS.qcow2 > > guestfs-sinmkv97r4k4dzcd_VARS.qcow2 guestfs-vj5cu9hgwcv0f18u_VARS.qcow2 > > guestfs-6jnim67ceme3h369_VARS.qcow2 guestfs-89g707uzdttxb9hv_VARS.qcow2 > > guestfs-h6vzt9zqxhoxx0hj_VARS.qcow2 guestfs-k4hw4gkcq8u0eebj_VARS.qcow2 > > guestfs-pljksgeblxgyf0pw_VARS.qcow2 guestfs-t8bnntykpzm3qvh2_VARS.qcow2 > > guestfs-ynbxqtj01sn4ovwy_VARS.qcow2 guestfs-6sqhenv6ditrq410_VARS.qcow2 > > guestfs-at8lndx32y69bjmg_VARS.qcow2 guestfs-ht0uycdn3vwdozad_VARS.qcow2 > > guestfs-lfy19ftph83jo4v0_VARS.qcow2 guestfs-qj0c2u9w2k4ors6h_VARS.qcow2 > > guestfs-tmq0h36spng4vtwk_VARS.qcow2 guestfs-7erqks7ks5c8rj38_VARS.qcow2 > > guestfs-ba2kzqxb1pna12l5_VARS.qcow2 guestfs-is7avsxjtnig5kgb_VARS.qcow2 > > guestfs-m3nrhs3x58rlget9_VARS.qcow2 guestfs-rilsled4ff9wvk6v_VARS.qcow2 > > guestfs-veebmlno2okc7c8h_VARS.qcow2 > > cel@boudin:~$ > > > > Or perhaps I do not understand what these files are. > > Based on the names, I believe those have been created by libguestfs. > > Rich, is it possible that libguestfs doesn't use > VIR_DOMAIN_UNDEFINE_NVRAM when cleaning up after itself? Libguestfs uses transient VMs, so there's no Undefine operation that will ever be invoked. We have a feature gap here in that we don't provide a mechanism for apps to say they want clean up of NVRAM and TPM state for transient VMs. We can't automatically clean up after transient VMs, as some will be intended to be restarted later. We already have VIR_DOMAIN_CREATE_AUTO_DESTORY which libguests uses to guarantee the VM is killed if libguest is killed. Again though, we can't assume that AUTO_DESTORY implies cleanup. Perhaps we need a VIR_DOMAIN_CREATE_AUTO_CLEANUP flag for virDomainCreateXML ? With regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :| _______________________________________________ Users mailing list -- users@xxxxxxxxxxxxxxxxx To unsubscribe send an email to users-leave@xxxxxxxxxxxxxxxxx
