When the metadata doesn't identify which array a spare belongs to
we normally require an explicit domain match to connect a spare
with an array.
However when the spare is explicitly listed in argv, it should be
safe to include as long as there is no domain conflict.
Signed-off-by: NeilBrown <neilb@suse.de>
Sometime we will need to know the difference between no domains found
and domains didn't match.
So allow domain_test to return different values and fix up all callers
to maintain current behaviour.
Signed-off-by: NeilBrown <neilb@suse.de>
For containers, it is always appropriate to include a device in the
container.
Whether it should then be included in an array is a separate question.
Signed-off-by: NeilBrown <neilb@suse.de>
If we find a device that has not superblock, we currently fail
unless in auto_assem mode.
However we really should only fail if the device was explicitly listed
in the arg list. So add a test for that.
Signed-off-by: NeilBrown <neilb@suse.de>
When there are any arrays in config file the spares with
domain not matching any array are not assembled because
auto assembly is not attempted.
Addition of ARRAY line with uuid=0:0:0:0 in config will work
with modified condition for gathering spares.
Signed-off-by: Anna Czarnowska <anna.czarnowska@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Arrays on partitions are not supported for external metadata
so do not take such spare from native array.
Signed-off-by: Anna Czarnowska <anna.czarnowska@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Too big map was copied (outside allocated memory) and this causes
mdmon crash for 2 raid0 arrays in container.
Map of correct (smaller) size should be copied,
to not overwrite any internal data.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When expansion is run on 2 raid0 arrays in container no update
is sent to mdmon because mdmon is off (mdadm performs update)
Memory size for first reshaped array is allocated to satisfy memory
requirements for expanded maps.
Memory for second device is allocated using old disks number, as in
metadata there is no information about this array reshape.
When mdmon initiates second array reshape it overwrites internal
structures and crashes). There is no place to keep expanded maps.
To avoid this situation during loading metadata, allocated memory
should be performed using the maximum used disks number in particular
container.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When second array reshape is about to start external metadata should
be updated by mdmon in imsm_set_array_state(). For this purposes
imsm_progress_container_reshape() is reused.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Reshape is started from second array, so it causes imsm incompatibility
and problems during second array start.
Reshape should be started in arrays metadata order.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
reshape_prepare_fdlist and child_monitor currently have slightly
different ideas of the 'old number of raid devices' which can cause
major confusion.
So settle on one value, and assign it to odisks early and always use
it.
Signed-off-by: NeilBrown <neilb@suse.de>
kernel requires sync_max to be a multiple of the current
chunk size. This is not really 'correct', but we need to work
with it. So round down.
Signed-off-by: NeilBrown <neilb@suse.de>
Due to a miscalculation we didn't initialise the whole file.
There is 4K (8 sectors) for the metadata, then the data.
Signed-off-by: NeilBrown <neilb@suse.de>
When restarting a reshape with internal metadata, the new geometry
is already set and the reshape has been start (but has not been
allowed to continue yet).
So in that case, don't set things and don't ask for a reshape.
Signed-off-by: NeilBrown <neilb@suse.de>
The value of 'array' might not be current, so SET_ARRAY_INFO
and fail. Just refresh it before setting raid_disks.
Signed-off-by: NeilBrown <neilb@suse.de>
Now that write_init_super doesn't close fds any more, we need
to call free_super before the ADD_NEW_DISK ioctl.
Also call free_super before some error returns, for cleanliness.
Signed-off-by: NeilBrown <neilb@suse.de>
Includes description of POLICY line in /etc/mdadm.conf
and of changes in Monitor and Incremental related to autorebuild.
Signed-off-by: Anna Czarnowska <anna.czarnowska@intel.com>
Signed-off-by: Przemyslaw Czarnowski <przemyslaw.hawrylewicz.czarnowski@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Test 9: We do not block spare migration between different metadatas.
test 13: Migrated spare must belong the same domain as destination -
there is no additional condition for action.
Signed-off-by: Marcin Labun <marcin.labun@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
By default IMSM checks if member device belongs to AHCI or ISCI controller.
When using loop devices one must disable these checks by setting
IMSM_NO_PLATFORM.
Signed-off-by: Marcin Labun <marcin.labun@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
As free_super now closes fds for member devices, rather than
write_init_super doing it, we need to call free_super earlier,
so that the device (on which we hold an O_EXCL open) is closed
before it is added to the array.
So close at the end of pass-1 rather than after pass-2.
Signed-off-by: NeilBrown <neilb@suse.de>
Now that a 'supertype' container more information, the simplistic
copying of 'st' into 'refst' is incorrect and results in closing
some fds when load_super1(refst) calls free_super().
So do it more correctly using dup_super.
Reported-by: "Labun, Marcin" <Marcin.Labun@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Adding spare disks to imsm container fails due to problem with writing
new_dev to sysfs. This problem was caused by not closed handle
(opened exclusively) in Manage.c:803.
Disk handle was not closed by free_imsm().
This is due to not released disk_mgmt_list in free_imsm_disks().
Proper release of imsm metadata allows for spare adding without problems.
Memory leak was fixed also.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Tests revealed that sometimes there are still more spares taken
than needed. The reason for this is that after adding one spare
to container with degraded subarray
if between ioctl in main loop and load_container in try_spare_migration
mdmon activates the spare we see active<raid but find no spares in parent
container and so add an extra spare.
To prevent such behaviour we count active disks in the list returned
by getinfo_super_disks and compare it with subarray->active.
If the number has increased it means new spare was added and activated
so there is no need for more.
Signed-off-by: Anna Czarnowska <anna.czarnowska@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Problem has been observed when raid10<->raid0 takeover operation
is executed.
In code updating layout, raid_disks and chunk_size for non-restriping
operations in reshape array functions SET_ARRAY_INFO ioctl call was
not succeeded.
Takeover process finish execution with error, mdadm shows message:
"mdadm: failed to set disks"
Cause is not meeting SET_ARRAY_INFO ioctl requirements:
- only one parameter may be changed at one time
- level of current array info and new info should be the same
Patch introduces solution for this issue.
At the beginning of discussed code we read current information
about array and then compare them with new values should be set.
If particular value is different (and should be set),
we are overwrite only this one in array info and then call ioctl.
Signed-off-by: Krzysztof Wojcik <krzysztof.wojcik@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
For raid10 -> raid0 takeover operation we should reject disks
in mirror by marking them as 'failed' and then remove them from
array by writing "remove" to disk state.
For external metadata second action is executed by mdmon.
According the description in monitor.c:175 when monitor detect
"faulty" in device state, it blocks the device, mark it as failed
in metadata, unblocks the device and finally writes "remove"
to device state.
For external case writing "remove" to device state in mdadm
is not necessary and harmful.
It may cause following issues:
1. "remove" operation for external case in mdadm is not finish
with successful result because monitor may block the device or disk
has been already removed by monitor.
2. If disk is removed by mdadm earlier than mdmon catch "failed" state,
metadata is not properly updated- is not marked as failed.
Signed-off-by: Krzysztof Wojcik <krzysztof.wojcik@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
After loop can occurs that due to 0 value reported by kernel
we have 0 in completed variable.
This is wrong. we are interested in real completed point.
0 value means that we reached sync point set in md,
so we can set completed variable to just reached point.
this point value is stored in max_progress variable.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
mdadm should verify if reshape is started before it goes
in to check-pointing machine.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Problem occurs when we want to expand single disk raid0 array.
This is done via degraded 2 disks raid4 array. When new spare
is added to array, md immediately initiates recovery before
mdadm can configure and start reshape. This is due fact that 2 disk
raid4/5 array is special md case. Mdmon does nothing here because
container is blocked.
Put array in to frozen state allows mdadm to finish configuration
before reshape is executed in md.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
imsm_reshape_super() currently allows for expansion when requested
raid_disks number is the same as current.
This is wrong. Existing in code condition is too weak.
We should allow for expansion when new disks_number is greater
than current only.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
# mdadm --detail --export /dev/md127p1
Before:
MD_LEVEL=raid5
MD_DEVICES=4
MD_METADATA=0.90
After:
MD_LEVEL=raid5
MD_DEVICES=4
MD_CONTAINER=/dev/md0
MD_MEMBER=0
MD_UUID=55746a20:925d24a7:4f9bd7e2:9c9a411f
We parse the symlink target with a format:
../../block/mdXXX/mdXXXpYY
...and need the second '/' from the end of the string to read detect a
'md' device.
Reported-by: Krzysztof Wasilewski <krzysztof.wasilewski@intel.com>
Cc: Przemyslaw Czarnowski <przemyslaw.hawrylewicz.czarnowski@intel.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Neil,
Please consider this patch that once was discussed and I think agreed with in general direction. It was sent a while ago
but somehow did not merged into your devel3-2. This patch enables hot-plug of so called bare devices (as understand by domain policies rules in mdadm.conf).
Without this patch we do NOT serve hot-plug of bare devices at all.
Thanks,
Marcin Labun
Subject was: FW: Autorebuild, new dynamic udev rules for hot-plugs
>>From c0aecd4dd96691e8bfa6f2dc187261ec8bb2c5a2 Mon Sep 17 00:00:00 2001
From: Przemyslaw Czarnowski <przemyslaw.hawrylewicz.czarnowski@intel.com>
Date: Thu, 23 Dec 2010 16:35:01 +0100
Subject: [PATCH] Dynamic hot-plug udev rules for policies
Cc: linux-raid@vger.kernel.org, Williams, Dan J <dan.j.williams@intel.com>, Ciechanowski, Ed <ed.ciechanowski@intel.com>
When introducing policies, new hot-plug rules were added to support
bare disks. Mdadm was started for each hot plugged block device
to determine if it could be used as spare or as a replacement member for
degraded array.
This patch introduces limitation of range of devices that are handled
by mdadm.
It limits them to the ones specified in domains associated with
the actions: spare-same-port, spare and spare-force.
In order to enable hot-plug for bare disks one must update udev rules
with command
mdadm --activate-domains[=filename]
Above command writes udev rule configuration to stdout. If 'filename'
is given output is written to the file provided as parameter. It is up
to system administrator what should be done later. To make such rule
permanent (i.e. remain after reboot) rule should be writen to
/lib/udev/rules.d directory. Other cases will just need to write it to
/dev/.udev/rules.d directory where temporary rules lies. One should be
aware of the meaning of names/priorities of the udev rules.
After mdadm.conf is changed one is obliged to re-run
"mdadm --activate-domains" command in order to bring the system
configuration up to date.
All hot-plugged disks containing metadata are still handled by existing
rules.
Signed-off-by: Przemyslaw Czarnowski <przemyslaw.hawrylewicz.czarnowski@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
When analyse_change sets level=1, data_disks is meaningless
as is layout.
So don't set them, and make sure we ignore them.
Signed-off-by: NeilBrown <neilb@suse.de>
Add support for raid1 to raid0 takeover operation in user space.
This patch includes support for native and imsm metadata.
Signed-off-by: Krzysztof Wojcik <krzysztof.wojcik@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
During reshape when reshape is finished in md, progress_reshape() hangs
on select().
This is because 'sync_completed' is reset to zero before 'sync_action'
becomes 'idle', and we don't look for notification on 'sync_action'.
So if completed becomes zero after reshape_progress has made some
progress, then deduce that reshape has finished.
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
when in container are present raid0 and raid5 arrays, and reshape order is:
1. raid0 array
2. raid5 array
mdadm cannot set new raid_disks for raid0 array. For this action md has to have
handshake with mdmon. We have the following conditions:
1. Raid0 is not monitored
2. raid0 has been just takeovered to raid4/5 (it has to be monitored
3. monitor has to start monitor new raid4/5 array
4. monitor is not started (it is started to second raid5 array)
In such situation pig_monitor is required to let know to m monitor about new array
(not in the starting monitor case only)
Signed-off-by: Adam Kwolek <adam.kwolek@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
get_disk_controller_domain recognizes Intel (R) SAS controller (isci).
The function returns three different strings that differentiate disk attached
to AHCI, ISCI or unknown controller types to create separate domains
for each case.
Signed-off-by: Marcin Labun <marcin.labun@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Added support in detail_platform_imsm for Intel (R) SAS controller.
Function supports AHCI and ISCI controllers. RAID properties are derived
from common OROM for both types.
Based on code From: Artur Wojcik <artur.wojcik@intel.com>
Signed-off-by: Marcin Labun <marcin.labun@intel.com>
Signed-off-by: NeilBrown <neilb@suse.de>
NeilBrown
Czarnowska, Anna
Adam Kwolek
Przemyslaw Czarnowski
Labun, Marcin
Krzysztof Wojcik
Dan Williams