I don't think anyone can measure latency to 1% accuracy. It's better to not even pretend to be able to. 10% should be fine even for the most latency-critical environments. Signed-off-by: Martin Wilck <mwilck@xxxxxxxx> --- libmultipath/prioritizers/path_latency.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/libmultipath/prioritizers/path_latency.c b/libmultipath/prioritizers/path_latency.c index 9d5397ec1b3a..b8c5bc7c50a4 100644 --- a/libmultipath/prioritizers/path_latency.c +++ b/libmultipath/prioritizers/path_latency.c @@ -42,8 +42,9 @@ #define DEF_IO_NUM 100 #define MAX_BASE_NUM 10 -#define MIN_BASE_NUM 1.01 -#define DEF_BASE_NUM 1.5 +#define MIN_BASE_NUM 1.1 +// This is 10**(1/4). 4 prio steps correspond to a factor of 10. +#define DEF_BASE_NUM 1.77827941004 #define MAX_AVG_LATENCY 100000000. /* Unit: us */ #define MIN_AVG_LATENCY 1. /* Unit: us */ -- 2.15.0 -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
