[PATCH 3/7] drbd: raise the socket buffer size limits to 128 MiB

[Christoph Böhmwalder]

sndbuf-size and rcvbuf-size are capped at 10 MiB, with a comment that a
larger TCP buffer is not useful. That was written for the networks of
2008. On today's long-haul 25G+ links the bandwidth-delay product of a
single DRBD connection easily exceeds 10 MiB, and the cap then limits
replication throughput regardless of how the rest of the stack is tuned.

Raise both limits to 128 MiB, the value DRBD 9 has used for years. The
defaults are unchanged: 0 keeps the kernel's auto-tuning.

Signed-off-by: Christoph Böhmwalder <christoph.boehmwalder@xxxxxxxxxx>
---
 include/uapi/linux/drbd_limits.h | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

diff --git a/include/uapi/linux/drbd_limits.h b/include/uapi/linux/drbd_limits.h
index 65d3ebf2437b..4132b619376c 100644
--- a/include/uapi/linux/drbd_limits.h
+++ b/include/uapi/linux/drbd_limits.h
@@ -87,14 +87,13 @@
 #define DRBD_MAX_EPOCH_SIZE_DEF 2048U
 #define DRBD_MAX_EPOCH_SIZE_SCALE '1'
 
-  /* I don't think that a tcp send buffer of more than 10M is useful */
 #define DRBD_SNDBUF_SIZE_MIN  0U
-#define DRBD_SNDBUF_SIZE_MAX  (10U<<20)
+#define DRBD_SNDBUF_SIZE_MAX  (128U << 20)
 #define DRBD_SNDBUF_SIZE_DEF  0U
 #define DRBD_SNDBUF_SIZE_SCALE '1'
 
 #define DRBD_RCVBUF_SIZE_MIN  0U
-#define DRBD_RCVBUF_SIZE_MAX  (10U<<20)
+#define DRBD_RCVBUF_SIZE_MAX  (128U << 20)
 #define DRBD_RCVBUF_SIZE_DEF  0U
 #define DRBD_RCVBUF_SIZE_SCALE '1'
 
-- 
2.55.0