The following patch fixes an off-by-one error in ioc_conv(). Calls to
this function are corrupting device names (at least on IA-64), so
comparing the device name to the name in diskstats always fails. Hence
no disk output.
This patch will apply to both the upstream tarball and the debian
package
Package: multipath-tools
Version: 0.6.1-3
When using parted to create partitions, multipath partition device
files are created using kernel names rather than honoring the "-p
-part" option in the various udev rules. This issue exists in
multipath-tools/kpartx_0.6.1-3, but has been seen in earlier
to 256 LUNs (1024 paths)
with no problems.
I have seen similar issues in earlier versions of multipath-tools.
--
Andrew Patterson
Hewlett-Packard Enterprise
On 06/15/2016 08:43 AM, Ritesh Raj Sarraf wrote:
> Control: tag -1 moreinfo
>
> On Tue, 2016-06-14 at 17:30 -0600, Andrew Patterson wrote:
>> There is no output in dmesg.
>>
>> Changing /proc/sys/fs/aio-max-nr from the default 65536 to 1048576
>> seems to help.
On 06/15/2016 08:26 AM, Ritesh Raj Sarraf wrote:
> Thank you for testing the new version. FYI, I've pushed it to Unstable. So, in
> all probability, this is the version that'll be shipped with Stretch.
>
>
> On Tue, 2016-06-14 at 14:13 -0600, Andrew Patterson wrote:
&g
On Thu, 16 Jun 2016 00:15:37 +0530 Ritesh Raj Sarraf wrote:
> On Wed, 2016-06-15 at 11:12 -0600, Andrew Patterson wrote:
> >
> > Thanks for reminding me about max_fds. At boot:
> >
> > # cat /proc/sys/fs/nr_open
> > 1048576
> >
> > and
> >
module}."
else
log_failure_msg "failed to load module ${module}."
fi
done
All three of these methods should load the same set of modules (the
systemd one seems to be best). Perhaps, there should be a
/etc/default/multipath-tools config file that governs which modules to
load?
--
Andrew Patterson
Hewlett-Packard Enterprise
now getting events for the partition as well as the parent
block device.
Note that when using parted, we get a different partition device file
-- /dev/mapper/mpathb1 than that created with fdisk. This difference
is discussed in bug
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=827308/
I have seen this same issue in earlier versions of multipath-tools.
--
Andrew Patterson
Hewlett-Packard Enterprise
On Thu, 16 Jun 2016 15:28:42 +0530 Ritesh Raj Sarraf wrote:
> On Wed, 2016-06-15 at 14:05 -0600, Andrew Patterson wrote:
> > The systemd multpath-tools.service loads multipath modules using the
> > following
> > line:
> >
> > ExecStartPre=/sbin/modprobe -a scsi
> ../dm-4
lrwxrwxrwx 1 root root 7 Jun 16 10:29 mpathf -> ../dm-5
lrwxrwxrwx 1 root root 7 Jun 16 10:29 mpathg -> ../dm-6
lrwxrwxrwx 1 root root 7 Jun 16 10:29 mpathh -> ../dm-7
--
Andrew Patterson
Hewlett-Packard Enterprise
On Thu, 16 Jun 2016 22:54:08 +0530 Ritesh Raj Sarraf wrote:
> On Thu, 2016-06-16 at 22:47 +0530, Ritesh Raj Sarraf wrote:
> > On Thu, 2016-06-16 at 22:35 +0530, Ritesh Raj Sarraf wrote:
> > > Control: tag -1 +pending
> > >Â
> > > On Thu, 2016-06-16 at
reating link '/dev/disk/by-id/wwn-0xpatha-part1' to '/dev/dm-8'
preserve already existing symlink '/dev/disk/by-id/wwn-0xpatha-part1' to
'../../dm-8'
created db file '/run/udev/data/b254:8' for '/devices/virtual/block/dm-8'
.ID_FS_TYPE_NEW=
ACTION=remote
DEVLINKS=/dev/disk/by-id/wwn-0xpatha-part1 /dev/disk/by-id/scsi-mpatha-part1
DEVNAME=/dev/dm-8
DEVPATH=/devices/virtual/block/dm-8
DEVTYPE=disk
DM_MPATH=mpatha
DM_NAME=mpatha-part1
DM_PART=1
DM_STATE=ACTIVE
DM_TABLE_STATE=LIVE
DM_TYPE=scsi
DM_UUID=part1-mpath-360002ac0190428be
DM_WWN=0xpatha
ID_FS_TYPE=
MAJOR=254
MINOR=8
SUBSYSTEM=block
TAGS=:systemd:
USEC_INITIALIZED=363164888098
Unload module index
Unloaded link configuration context.
--
Andrew Patterson
Hewlett-Packard Enterprise
On Thu, 16 Jun 2016 15:19:22 +0530 Ritesh Raj Sarraf wrote:
> Control: tag -1 +pending
>
> On Wed, 2016-06-15 at 13:03 -0600, Andrew Patterson wrote:
> > For most users this is not an issue. How many systems are going to use
> > that many LUNs?
> >
>
> I agr
k there is
> a bug which processes '-' before '+' in version comparison...
>
> --
> Nicholas Luedtke
> Linux for HP Helion OpenStack, Hewlett-Packard Enterprise
Shouldn't this point to upstream rather than debian. In other words --
http://git.opensvc.com/gitweb.cgi?p=multipath-tools/.git;a=tags
Andrew
--
Andrew Patterson
Hewlett-Packard Enterprise
On Thu, 16 Jun 2016 22:13:10 +0530 Ritesh Raj Sarraf wrote:
> On Thu, 2016-06-16 at 10:20 -0600, Andrew Patterson wrote: > > > Yes. That'd
> be good. Perhaps the module loading could be
delegated to a > > separate > > > shell script. Such a script can then
be
ch-sonames libmultipath0
--
Andrew Patterson
Hewlett-Packard Enterprise
On Wed, 22 Jun 2016 18:44:49 + Niels Thykier wrote:
> On Wed, 22 Jun 2016 11:42:54 -0600 Andrew Patterson >
> wrote: > > Package: multipath-tools > >
Version: 0.6.1-3 > > > > Hi, > > Please keep me in CC for follow ups on
this if my reply is desire
On Wed, 22 Jun 2016 11:18:41 -0600 Andrew Patterson
wrote:
I forgot to add the multipath-tools.default file in the patch. Here is
the revised patch.
--
Andrew Patterson
Hewlett-Packard Enterprise
commit 566cc94c2fbcf1c59e5d3f8c30fd08c567f8a382
Author: Andrew Patterson
Date: Wed Jun 22 10
On Mon, 20 Jun 2016 16:40:29 -0600 Andrew Patterson
wrote:
> On Thu, 16 Jun 2016 15:19:22 +0530 Ritesh Raj Sarraf wrote:
> > Control: tag -1 +pending
> >
> > On Wed, 2016-06-15 at 13:03 -0600, Andrew Patterson wrote:
> > > For most users this is not an issue. How
On 06/25/2016 02:15 AM, Ritesh Raj Sarraf wrote:
> On Wed, 2016-06-22 at 16:08 -0600, Andrew Patterson wrote: >> if [ -f
> /etc/multipath-tools-boot.conf ]; then >> +.
/etc/multipath-tools-boot.conf >> +else >> +
MULTIPATH_TOOLS_HW_HANDLERS="" >
On 06/25/2016 02:16 AM, Ritesh Raj Sarraf wrote:
> On Thu, 2016-06-23 at 16:50 -0600, Andrew Patterson wrote: >> +# Save boot
> setting for aio-max-nr. >> +BOOT_AIO_MAX_NR=$(cat
/proc/sys/fs/aio-max-nr) >> + >> if [ -f /etc/multipath-tools-boot.conf
]; then
tags + patch
Here's a fix for the GCC 5 build issue. It adds 'export CFLAGS =
-std=gnu89' to the rules file to use the GNU89 inline semantics.
I tested it with both gcc-5.0 and gcc-4.9.
--
Andrew Patterson
Hewlett-Packard
--- padevchooser-0.9.4/debian/rules.orig 2
tags 778055 + patch
Here's a fix for the GCC 5 build issue. It adds 'export CFLAGS =
-std=gnu89' to the rules file to use the GNU89 inline semantics.
I tested it with both gcc-5.0 and gcc-4.9.
--
Andrew Patterson
Hewlett-Packard
--- padevchooser-0.9.4/debian/rules.orig 2
tags 778055 + patch
thanks
Here's a fix for the GCC 5 build issue. It adds 'export CFLAGS =
-std=gnu89' to the rules file to use the GNU89 inline semantics.
I tested it with both gcc-5.0 and gcc-4.9.
--
Andrew Patterson
Hewlett-Packard
--- padevchooser-0.9.4/debian/rules.ori
tags 778056 + patch
thanks
Here's a fix for the GCC 5 build issue. It adds 'export CFLAGS =
-std=gnu89' to the rules file to use the GNU89 inline semantics.
I tested it with both gcc-5.0 and gcc-4.9.
Note: we still get a lot of compiler warnings, but that is not a gcc-5 iss
Package: openvpn-auth-ldap
Version: 2.0.3-5.1
Tags: wheezy
The openvpn-auth plugin gets a segmentation fault at startup. This can
be demonstrated with the following command:
# openvpn --dev null --plugin /usr/lib/openvpn/openvpn-auth-ldap.so
Sat Oct 4 16:59:56 2014 OpenVPN 2.2.1 x86_64-linux-gnu
On Wed, 8 Oct 2014 18:40:40 +0200 Alberto Gonzalez Iniesta
wrote:
> Hi Andrew,
>
> The plugin segfaults because it needs a config file as a parameter.
> It's not a nice behaviour, but not a critial one (since it needs a
> config file to be useful).
>
> Try:
>
> openvpn --dev null --plugin /usr/li
his option always picks the appropriate
device. Example:
grub-installer/bootdev string default
Here is the patch:
Author: Andrew Patterson
Date: Thu Aug 14 14:00:47 2014 -0600
Using "default" for bootdev will install grub on default disk.
The grub-install/bootdev conf
Package: apt-cache-ng
Version: 0.7.11-2
Severity: important
Tags: wheezy
I get the following error when trying to upgrade apt-cache-ng on wheezy:
Setting up apt-cacher-ng (0.7.11-2) ...
adduser: The user `apt-cacher-ng' already exists. Exiting.
dpkg: error processing apt-cacher-ng (--configure):
On Sun, 31 Aug 2014 21:32:01 +0200 Eduard Bloch wrote:
> Control: forcemerge 706382 759740
>
> > > I get the following error when trying to upgrade apt-cache-ng on wheezy:
> > >
> > > Setting up apt-cacher-ng (0.7.11-2) ...
> > > adduser: The user `apt-cacher-ng' already exists. Exiting.
> > > dpk
On Wed, 3 Sep 2014 19:13:46 +0200 Eduard Bloch wrote:
> reassign 759740 adduser
> retitle 759740 --system --quiet ... still produces warnings and bad exit code
> thanks
>
> Hallo,
> * Andrew Patterson [Tue, Sep 02 2014, 03:24:10PM]:
>
> > I guess I don't unde
>
> I don't know. And as I mentioned before, feel free to use the backport
> version. It contains a quick workaround for a similar situation:
>
># user should exist. adduser sometimes fails (system range issue) but
> that's ok
>adduser --quiet --system --group --no-create-home --home "$CDI
covery/setup to fw_get_targets()
* Fix broken long command-line options in iscsiuio.
I have tested these patches using bnx2x devices configured as both
software initiators and using hardware offload.
--
Andrew Patterson
Hewlett-Packard Enterprise
>From 17b79c2a25a9ba5a105661f49bf435148ff30277 M
.
Setting up software interface failed
A subsequent run of iscsistart -f and iscsistart -b works fine. The
iscsstart -N routines should probable bypass be2iscsi devices given
that Emulex driver/firmware sets up networking on it own.
--
Andrew Patterson Hewlett-Packard Enterprise
uot;cxgb4i" },
{"bnx2", "bnx2i" },
{"bnx2x", "bnx2i"},
{NULL, NULL}
};
to see if -N should skip this NIC.
There is a discussion of this issue at:
https://groups.google.com/forum/?_escaped_fragment_=msg/open-iscsi/PsT65Z4G
failed boot.
Some discussion for this issue can be found at:
https://groups.google.com/forum/?_escaped_fragment_=topic/open-iscsi/PsT65Z4Gx3I#!topic/open-iscsi/PsT65Z4Gx3I
--
Andrew Patterson
Hewlett-Packard Enterprise
On 01/03/2017 01:02 PM, Christian Seiler wrote:
> Control: tags -1 + confirmed moreinfo
> Control: severity -1 wishlist
>
> On 01/03/2017 06:02 PM, Andrew Patterson wrote:
>> Version 2.0.874 of the open-iscsi package requires that the iscsiuio
>> daemon to be running befor
On 01/03/2017 01:27 PM, Christian Seiler wrote:
> On 01/03/2017 09:10 PM, Andrew Patterson wrote:
>>> How does the rest of the boot process proceed then? What happens
>>> when iscsiuio is to be started regularly at boot from the systemd
>>> service / init script
On 01/03/2017 03:47 PM, Christian Seiler wrote:
> On 01/03/2017 10:51 PM, Andrew Patterson wrote:
>> On 01/03/2017 01:27 PM, Christian Seiler wrote:
>>> On 01/03/2017 09:10 PM, Andrew Patterson wrote:
>>>>> How does the rest of the boot process proceed then? What
ISCSISTART_FLAGS="-S"
fi
.
.
.
modprobe iscsi_ibft
iscsistart -N $ISCSISTART_FLAGS
The iscsistart code would have to be modified to add support for this
flag. Alternatively, iscsistart could check for the presence of
/sbin/iscsiuio and "do the right thing".
--
Andrew Patterson
Hewlett-Packard Enterprise
On 01/04/2017 04:11 PM, Christian Seiler wrote:
> Control: reopen -1
>
> Hi,
> (reopening the bug report since iscsiuio doesn't actually work
> according to what you're telling me)
>
> On 01/04/2017 11:30 PM, Andrew Patterson wrote:
>> 1. In debian/extra/in
On 01/05/2017 11:10 AM, Andrew Patterson wrote:
>
> On 01/04/2017 04:11 PM, Christian Seiler wrote:
>> Control: reopen -1
>>
>> Hi,
>> (reopening the bug report since iscsiuio doesn't actually work
>> according to what you're telling me)
>&g
On 01/05/2017 12:10 PM, Christian Seiler wrote:
> On 01/05/2017 07:10 PM, Andrew Patterson wrote:
>> On 01/04/2017 04:11 PM, Christian Seiler wrote:
>>> OTOH, initramfs should write to /run/initramfs only, so maybe
>>> we should pass -p /run/initramfs/iscsiuio.pid to i
to what the symlink would
point to. But we are seeing other issue. But we are seeing other
issues like device files reappearing after the running multipath -f on
it, removing the LUN, and then running multipath -r which we think is
related.
--
Andrew Patterson
Hewlett-Packard Enterprise
On Fri, 15 Apr 2016 16:07:48 -0600 Andrew Patterson
wrote:
> On Sat, 16 Apr 2016 00:42:33 +0530 Ritesh Raj Sarraf wrote:
> > > On Sat, 2016-04-16 at 00:35 +0530, Ritesh Raj Sarraf wrote: > >
Hello Mike, > > > > All I can say is that I am able to see the >
b
that file
generates the blacklist:
kernels="$( (echo "$1
$unamer"; for deb in $debkernels; do echo "$list" | awk "\$2 == \"$deb\" {
print \$1; }"; done; ) \
| sed -e 's#\.#\\.#g' -e '/^$/ d' | sort -u)"
The sed line is
On 07/06/2016 02:04 PM, Julian Andres Klode wrote:
> On Wed, Jul 06, 2016 at 01:48:17PM -0600, Andrew Patterson wrote:
>> Package: apt
>> Version: 1.2.11
>>
>> When I run apt-get autoremove, apt will also remove kernels with a "+"
>> in the package na
On 07/06/2016 02:38 PM, Andrew Patterson wrote:
>
>
> On 07/06/2016 02:04 PM, Julian Andres Klode wrote:
>> On Wed, Jul 06, 2016 at 01:48:17PM -0600, Andrew Patterson wrote:
>>> Package: apt
>>> Version: 1.2.11
>>>
>>> When I run apt-get autorem
I have a fix for this in spew-1.0.7. Which can be found at:
ftp://ftp.berlios.de/pub/spew/1.0.7/index.html
I had to add includes to handle the normal strlen, memcpy,
stuff.
Tested with gcc-4.3.
--
Andrew Patterson
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of
49 matches
Mail list logo
