It looks like partitions are being reported correctly, just not the main
device
$ cat /proc/diskstats | awk '{print $3, $7, $11, $7+$11, $13, $14}' | grep nvme
nvme0n1 22988 457568 480556 2440737408 2441207708
nvme0n1p1 22988 339624 362612 528884 978496
nvme0n1p2 0 0 0 0 0
nvme0n1p5 0 0 0 4 4
No
drivers/md/dm-stats.c:
static void dm_stat_round(struct dm_stat *s, struct dm_stat_shared *shared,
struct dm_stat_percpu *p)
{
/*
* This is racy, but so is part_round_stats_single.
*/
unsigned long long now, difference;
unsigned i
https://www.redhat.com/archives/dm-devel/2015-June/msg00042.html
https://www.redhat.com/archives/dm-devel/2015-June/msg00057.html
Something is really really screwy here.
Some stats are in ns, some are in ms:
Index: linux-4.1-rc7/Documentation/device-mapper/statistics.txt
===
iostat is showing the same thing btw.
See also https://access.redhat.com/solutions/1462993
--
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/1790181
Title:
/proc/diskstats shows weird valu
Here is my patch for munin diskstats:
diff -u diskstats-dist diskstats
--- diskstats-dist 2018-09-29 16:28:39.933727540 -0700
+++ diskstats 2018-11-12 09:54:56.850815309 -0800
@@ -234,6 +234,9 @@
# a given second, the device is nearly 100% saturated.
my $utilization = $tot_ticks /
https://forums.fedoraforum.org/showthread.php?318580-iostat-svctm-and-
busy-numbers-are-wrong-for-NVME-drives
I think summing wr_ticks and rd_ticks might actually be the right
solution.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linu
After rebooting, the values are < 2^31 and utilization is reporting
fine. I feel this is a 31/32 bit overflow issue.
--
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/1790181
Title:
/proc/d
7 matches
Mail list logo
