Skip to Main Content

Infrastructure Software

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

BUG: running parted causes reset of block device parameter

Peter_L_Aug 3 2017 — edited Aug 7 2017

Hi,

I'm using Oracle Linux 7.3. with NetApp LUNs and multipathing.

The NetApp LUNS has a block size of 4096, so the "max_sectors_kb" in OS has to set  to 4096, too.

If you use parted to check disk alignment or other things there is a major bug:

Important parameter "max_sectors_kb" is reset to 64 if you run and quit the parted program:

#cat /sys/class/block/sda/queue/max_sectors_kb

4096

# parted

GNU Parted 3.1

Verwende /dev/sda

Willkommen zu GNU Parted! Geben Sie 'help' ein, um eine Liste der verfügbaren Kommados zu erhalten.

(parted) quit

# cat /sys/class/block/sda/queue/max_sectors_kb

64

The reset of this value causes many errors like this:

kernel: blk_cloned_rq_check_limits: over max size limit.

Then you wil get flapping multipaths and the instance will crash, only resolvable by reboot!

Its become more important, because the orachk uses parted to test disks. During runing orachk the issue is occured in my env.

UPDATE:

Found this on Red Hat

https://access.redhat.com/solutions/43861

Have to add an udev rule:

#vi /etc/udev/rules.d/71-block-max-sectors.rules

ACTION=="add|change", SUBSYSTEM=="block", ENV{ID_VENDOR}=="NETAPP", RUN+="/bin/sh -c '/bin/echo 4096 >> /sys%p/queue/max_sectors_kb'"

Now it seems to work!

Could somebody inform the Oracle Linux devteam. I have no contact for doing so, thank you.

Regards Peter

Comments
Locked Post
New comments cannot be posted to this locked post.
Post Details
Locked on Aug 31 2017
Added on Aug 3 2017
0 comments
836 views