debian/patches/pve/0007-PVE-Config-rbd-block-rbd-disable-rbd_cache_writethro.patch

   1 From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
   2 From: Wolfgang Bumiller <w.bumiller@proxmox.com>
   3 Date: Mon, 6 Apr 2020 12:16:36 +0200
   4 Subject: [PATCH] PVE: [Config] rbd: block: rbd: disable
   5  rbd_cache_writethrough_until_flush
   6
   7 Either the cache mode asks for a cache or not. There's no
   8 point in having a "temporary" cache mode. This option AFAIK
   9 was introduced as a hack for ancient virtio drivers. If
  10 anything, we should have a separate option for it. Better
  11 yet, VMs affected by the related issue should simply
  12 explicitly choose writethrough.
  13
  14 Signed-off-by: Wolfgang Bumiller <w.bumiller@proxmox.com>
  15 Signed-off-by: Thomas Lamprecht <t.lamprecht@proxmox.com>
  16 ---
  17  block/rbd.c | 2 ++
  18  1 file changed, 2 insertions(+)
  19
  20 diff --git a/block/rbd.c b/block/rbd.c
  21 index e637639a07..5717e7258c 100644
  22 --- a/block/rbd.c
  23 +++ b/block/rbd.c
  24 @@ -651,6 +651,8 @@ static int qemu_rbd_connect(rados_t *cluster, rados_ioctx_t *io_ctx,
  25          rados_conf_set(*cluster, "rbd_cache", "false");
  26      }
  27
  28 +    rados_conf_set(*cluster, "rbd_cache_writethrough_until_flush", "false");
  29 +
  30      r = rados_connect(*cluster);
  31      if (r < 0) {
  32          error_setg_errno(errp, -r, "error connecting");