Currently if a md raid0/linear array gets one or more members removed while
being mounted, kernel keeps showing state 'clean' in the 'array_state'
sysfs attribute. Despite udev signaling the member device is gone, 'mdadm'
cannot issue the STOP_ARRAY ioctl successfully, given the array is mounted.
Nothing else hints that something is wrong (except that the removed devices
don't show properly in the output of mdadm 'detail' command). There is no
other property to be checked, and if user is not performing reads/writes
to the array, even kernel log is quiet and doesn't give a clue about the
missing member.
This patch is the mdadm counterpart of kernel new array state 'broken'.
The 'broken' state mimics the state 'clean' in every aspect, being useful
only to distinguish if an array has some member missing. All necessary
paths in mdadm were changed to deal with 'broken' state, and in case the
tool runs in a kernel that is not updated, it'll work normally, i.e., it
doesn't require the 'broken' state in order to work.
Also, this patch changes the way the array state is showed in the 'detail'
command (for raid0/linear only) - now it takes the 'array_state' sysfs
attribute into account instead of only rely in the MD_SB_CLEAN flag.
Cc: Jes Sorensen <jes.sorensen@gmail.com>
Cc: NeilBrown <neilb@suse.de>
Cc: Song Liu <songliubraving@fb.com>
Signed-off-by: Guilherme G. Piccoli <gpiccoli@canonical.com>
Signed-off-by: Jes Sorensen <jsorensen@fb.com>
After switch root new mdmon is started. It sends initrd mdmon a signal
to terminate. initrd mdmon receives it and switches the safe mode delay
to 1 ms in order to get array to clean state and flush last version of
metadata. The problem is sysfs filesystem is not available to initrd mdmon
after switch root so the original safe mode delay is unchanged. The delay
is set to few seconds - if there is a lot of traffic on the filesystem,
initrd mdmon doesn't terminate for a long time (no clean state). There
are 2 instances of mdmon. initrd mdmon flushes metadata when array goes
to clean state but this metadata might be already outdated.
Use file descriptor obtained on mdmon start to change safe mode delay.
Signed-off-by: Tomasz Majchrzak <tomasz.majchrzak@intel.com>
Signed-off-by: Jes Sorensen <jsorensen@fb.com>
The clang compiler complained about each of these.
The mdmon.h error will only affect 'far' RAID10 arrays using intel or DDF
metadata, and there is no such thing.
The mdopen.c will cause a problem if there are no free md device
numbers in the first 512. That is fairly unlikely.
The restripe.c error would only affect the 'test_stripe' command, and
probably doesn't change its behaviour.
The super-intel.c fix is purely cosmetic.
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Jes Sorensen <Jes.Sorensen@redhat.com>
Sometimes mdadm prints messages with wrong name "mdmon",
and vice versa.
This patch solves this problem by changing method of determining
process name.
Now "Name" will be set in const at start of a program,
previously was hardcoded as #define.
Signed-off-by: Pawel Baldysiak <pawel.baldysiak@intel.com>
Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
For RAID10, 'sync' numbers go up to the array size rather than the
component size. is_resync_complete() needs to allow for this.
Reported-by: Pawel Baldysiak <pawel.baldysiak@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
We widely use a "devnum" which is 0 or +ve for md%d devices
and -ve for md_d%d devices.
But I want to be able to use md_%s device names.
So get rid of devnum (a number) and use devnm (a 32char string).
eg.
md0
md_d2
md_home
Signed-off-by: NeilBrown <neilb@suse.de>
Description of the bug:
Sometimes mdmon crashes after changing RAID level from 1 to 0 (takeover).
Cause of the bug:
The managemon marks an active_array for removal from monitoring
by assigning a->container to NULL value (in the "manage_member" function).
Sometimes (during stress test) it happens right when the monitor
is in the "read_and_act" function and a->container pointer is in use.
This causes the monitor crashes.
Solution:
The active array has to be marked for removal in another way
than setting NULL pointer when it can be in use.
A new field "to_remove" was added to the "active_array" structure.
It is used in the managemon to mark a container to remove
(instead of the old assigment: a->container = NULL)
and monitor checks it to determine if the array should be removed.
The field "to_remove" should be checked in some other places
to avoid managing of the array which is going to be removed.
Signed-off-by: Lukasz Dorau <lukasz.dorau@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When mdadm starts a reshape, it might add some devices to the array
first. mdmon needs to notice the reshape starting and check for any
new devices. If there are any they need to be provided to be
monitored.
Signed-off-by: NeilBrown <neilb@suse.de>
The kernel updates and notifies md/sync_completed when it is time to
take a checkpoint. When this occurs (at 1/16 array size intervals)
write 'idle' to md/sync_action to have the current recovery position
updated in recovery_start and resync_start.
Requires the metadata handler to reset ->last_checkpoint when it has
determined that recovery has ended.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Now that we don't "mdadm --takeover" until /var/run is writable
there is no need to continually try to create files in there.
So only create these files at startup and fail if they cannot be
made. This means that to start an array with externally managed
metadata, either /var/run or ALT_RUN (e.g. /lib/init/rw) must be
writable. To 'takeover' from a previous mdmon instance, /var/run
must be writable.
This means we don't need to worry about SIGHUP (which was once used to
tell us it was time to create .pid) and SIGALRM.
Signed-off-by: NeilBrown <neilb@suse.de>
/var/run probably doesn't persist from early boot.
So if necessary, store in in /lib/init/rw or somewhere else
that does persist.
Signed-off-by: NeilBrown <neilb@suse.de>
We don't need to sprinkle reads of this attribute all over the place,
just once at the entry of read_and_act(). Also, the mdinfo structure
for the array already has a 'resync_start' member, so just reuse that.
Finally, rename get_resync_start() to read_resync_start to make it
consistent with the other sysfs accessors in monitor.c.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Now that mdmon handles sigterm if another monitor wants to take over it
should wait until all managed arrays are clean. So make WaitClean()
available to mdmon and teach try_kill_monitor() to wait on each subarray
in the container.
...since we may be communicating with a dieing process, we need to
block SIGPIPE earlier.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
We generally don't want mdmon to be terminated, but if a SIGTERM gets
through try to leave the monitored arrays in a clean state, block
attempts to mark the array dirty, and stop servicing the socket.
When we are killed by sigterm don't remove the pidfile let that be
cleaned up by the next monitor.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
md/resync_start reports different terminal values depending on kernel
configuration (~0UL versus ~0ULL). Make detection of the
resync-complete state more robust by comparing against array size.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
When arrays do not startup correctly it would be nice to know why. Need
to move the dprintf definition to mdadm.h
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Allow mdmon to start while /var/run/mdadm is readonly. Later a SIGHUP
can trigger mdmon to drop its pid and socket once /var/run/mdadm is
writable. Of course one needs the pid to send a HUP, that can be stored
in a distribution specific rw-init directory... For now, rely on a
killall -HUP mdmon to get the files dumped.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
If the metadata_version is
-mdXXX/whatever
rather than
/mdXXX/whatever
then the array is readonly and should be left alone by mdmon.
Signed-off-by: NeilBrown <neilb@suse.de>
When a 'ping' (empty message) is sent to mdmon, we wait for
'monitor' to do a full loop to make sure it has caught up
with anything that needs doing.
This allows synchronisation between mdadm and mdmon.
Maybe monitor should signal managemon rather than managemon polling...
Signed-off-by: Neil Brown <neilb@suse.de>
Only one superswitch should be externally visible for each
general type. Others which handle different flavours
(e.g. container/data-array) should be internal only.
for development only as console output can block leading to monitor deadlocks
in low mem situations
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Code in manager can now just call queue_metadata_update with a
(freeable) buf holding the update, and it will get passed to the
monitor and written out.
"sync_complete" just tracks the current resync/recover/check/whatever pass.
"resync_start" tracks which parts of the array are known to be in-sync
(modulo active writes). So it is what we need to use to update the metadata.
Also we cannot call it when the array has stopped, as the value is no longer
available then. We must call it when the resync completes.
Possibly also call it preiodically if the array is quiescent.
From: Dan Williams <dan.j.williams@intel.com>
Each md_message encapsulates a single command. A command includes an 'action'
member which describes what if any data comes after the action. Communication
with the monitor involves updating the active_cmd pointer and then writing to
mgr_pipe. Pass/fail status is returned via mon_pipe.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
From: Dan Williams <dan.j.williams@intel.com>
mdadm handles setting resync_start, monitor uses this value to determine
whether to set the 'active' or 'readauto' state.
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Guilherme G. Piccoli
Tomasz Majchrzak
NeilBrown
Pawel Baldysiak
NeilBrown
Lukasz Dorau
Dan Williams