similar to: [PATCH 0/5] Improve LVM handling in the appliance

Hi, this series improves the way LVM is used in the appliance: in particular, now lvmetad can eventually run at all, and with the correct configuration. Also improve the listing strategies. Changes in v2: - dropped patch #5, will be sent separately - move lvmetad statup in own function (patch #2) Thanks, Pino Toscano (4): daemon: lvm-filter: set also global_filter daemon: lvm-filter:

Simplify how we handle lvm.conf.

This is a simpler patch that removes GUESTFSD_EXT_CMD completely.

Currently lvmetad is started in init, and thus using the system (= appliance) configuration of lvm. Later on, in the daemon, a local copy of the lvm configuration is setup, and set it for use using the LVM_SYSTEM_DIR environment variable: this means only the programmes executed by the daemon will use the local lvm configuration, and not lvmetad. Thus manually start lvmetad from the daemon, right

Replace GUESTFSD_EXT_CMD with a command line option ‘./guestfsd --print-external-commands’

Hi, newer lvm2 releases don't have have uncommented "filter" lines, so the current way to edit lvm.conf doesn't work anymore. Instead, switch to augeas (with a "custom" len) for a cleaner and working way to set the lvm filter. Pino Toscano (3): daemon: add add_sprintf daemon: move AUGEAS_ERROR to the common header daemon: lvm-filter: use augeas for setting the

This patch reworks how we start up LVM and lvmetad. It fixes the problem we had converting a guest which had a peculiar LVM configuration: https://bugzilla.redhat.com/show_bug.cgi?id=1581810#c14 However please note I have NOT yet tested it fully. Rich.

When the daemon starts up it creates a fresh (empty) LVM configuration and starts up lvmetad (which depends on the LVM configuration). However this appears to cause problems: Some types of PV seem to require lvmetad and don't work without it (https://bugzilla.redhat.com/show_bug.cgi?id=1581810). If we don't start lvmetad earlier, the device nodes are not created. Therefore move the

lvm2 2.02.107 adds the -S/--select option used in lvs to filter out only public LVs (see RHBZ#1278878). To make this work again with versions of lvm2 older than that, only on old versions filter out thin layouts and compose the resulting device strings ourselves. The filtering done is much simplier than what "-S lv_role=public" will do, but should be good enough for our need. ---

On Wednesday, 18 July 2018 15:37:24 CEST Richard W.M. Jones wrote: > The old vgscan API literally ran vgscan. When we switched to using > lvmetad (in commit dd162d2cd56a2ecf4bcd40a7f463940eaac875b8) this > stopped working because lvmetad now ignores plain vgscan commands > without the --cache option. > > We documented that vgscan would rescan PVs, VGs and LVs, but without >

[This email is either empty or too large to be displayed at this time]

When a disk image uses LVM thinp (thin provisioning), the guestfs_lvs API would return the thinp pools. This confused other APIs because thinp pools don't have corresponding /dev/VG/LV device nodes. Filter the LVs that are returned using "lv_role=public". Thanks: Fabian Deutsch --- daemon/lvm.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/daemon/lvm.c

Don't carry around Augeas lenses. It is fragile, since if the lens is added to upstream Augeas but the version number has not changed, then Augeas won't parse the target file at all. This specifically causes password adjustments to fail in RHEL 7.1. In future, if we need an Augeas lens, it must be added to Augeas, either upstream or as a downstream patch carried around by distros. ---

guestfsd calls many different tools. Keeping track of all of them is error prone. This patch introduces a new helper macro to put the command string into its own ELF section: GUESTFSD_EXT_CMD(C_variable, command_name); This syntax makes it still possible to grep for used command names. The actual usage of the collected list could be like this: objcopy -j .guestfsd_ext_cmds -O binary

v2: - Changes as suggested by Pino in previous review.

v2 of previous patch, which fixes some missing bits. For now I'm going to go with Pino's RHEL 7.1 patch, since it is at least smaller than this. So I'm sending this to the list just to have it archived for later. Rich.

GUESTFSD_EXT_CMD was used by OpenSUSE to track which external commands are run by the daemon and package those commands into the appliance. It is no longer used by recent SUSE builds, so remove it. Thanks: Pino Toscano, Olaf Hering. --- daemon/9p.c | 3 +- daemon/available.c | 7 +-- daemon/base64.c | 6 +-- daemon/blkid.c | 10 ++---

v3: - Don't remove the LVM transform.

When listing logical volumes, ignore the ones which don't get activated automatically. No /dev/VG/LV device node is created for these ones which confuses APIs that attempt to do 'guestfs_lvs' followed by opening the device node. Note that 'guestfs_lvs_full' is unaffected by this change. --- daemon/lvm.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git

Hi, currently /etc/shadow is edited manually when needed (i.e. when setting the password for an user), and it is not changed when removing users. Import the upstream shadow.aug (currently in their development serie, but not part of any released version yet), and use it only when the augeas version is less than a potential 1.2.1 (covering also the case when the new version is just 1.3.0). Pino