This bug was fixed in the package linux - 5.0.0-7.8
---------------
linux (5.0.0-7.8) disco; urgency=medium
* linux: 5.0.0-7.8 -proposed tracker (LP: #1818519)
* Packaging resync (LP: #1786013)
- [Packaging] update helper scripts
* unnecessary request_queue freeze (LP: #1815733)
- block: avoid setting nr_requests to current value
- block: avoid setting none scheduler if it's already none
* Miscellaneous Ubuntu changes
- SAUCE: selftests: net: Don't fail test_vxlan_under_vrf on xfail
- update dkms package versions
[ Upstream Kernel Changes ]
* Rebase to v5.0
-- Seth Forshee <[email protected]> Mon, 04 Mar 2019 08:46:10
-0600
** Changed in: linux (Ubuntu Disco)
Status: Fix Committed => Fix Released
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1815733
Title:
unnecessary request_queue freeze
Status in linux package in Ubuntu:
Fix Released
Status in linux source package in Bionic:
New
Status in linux source package in Cosmic:
New
Status in linux source package in Disco:
Fix Released
Bug description:
Hi all,
In linux-4.15, which is used as hwe-edge in xenial, there's an unexpected
behavior. It results in latency spikes of generic_make_request() function call.
How to reproduce:
1. Run fio:
fio --name=test --rw=randwrite --direct=1 --filename=/dev/sdt --iodepth=1
--fsync=1 --ioengine=libaio --write_lat_log loglat
2. While fio is writing to the device, change nr_requests setting to the
same value as current:
echo 256 > /sys/block/sdt/queue/nr_requests
or: if scheduler is already none:
echo none > /sys/block/sdt/queue/scheduler
3. Stop fio and check slat log:
...
3980, 16, 1, 4096
3999, 17, 1, 4096
4010, 19, 1, 4096
4020, 18, 1, 4096
4045, 10821, 1, 4096
4054, 25, 1, 4096
4070, 12, 1, 4096
4083, 17, 1, 4096
4094, 16, 1, 4096
4111, 17, 1, 4096
...
There're a few patches to fix this behavior:
https://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-block.git/commit/?h=for-5.1/block&id=e5fa81408fb43ebabde65938ef8b20ae879017e7
https://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-block.git/commit/?h=for-5.1/block&id=fbd72127c975dc8e532ecc73d52f3b1b00935bec
This behavior affects production systems, because every change uevent
triggers udev to change settings prior to udev rules. Some software,
for example arcconf (tool for managing adaptec HBA adapters) triggers
uevents for every block device. So, every call to arcconf utility
results in latency spike.
Could this be backported to 4.15 kernel for xenial/bionic?
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1815733/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : [email protected]
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
