autofs cifs mount(s) hang / become momentarily unreachable

Question

I use autofs on my Fedora 25 desktop in order to access some cifs mounts on my work network.

/etc/auto.master:

/home/user/work/mountpoint /etc/auto.custom --timeout=43200 --ghost

/etc/auto.custom:

cifs1   -fstype=cifs,nodev,nosuid,uid=user,gid=user,credentials=/etc/.smbcredentials    ://server/cifs1_share/
cifs2   -fstype=cifs,nodev,nosuid,uid=user,gid=user,credentials=/etc/.smbcredentials    ://server/cifs2_share/
cifs3   -fstype=cifs,nodev,nosuid,uid=user,gid=user,credentials=/etc/.smbcredentials    ://server/cifs3_share/

This exact setup worked perfectly for a really long time:

Access to /home/user/work/mountpoint/{cifs1,cifs2,cifs3} was always fast, never hanged or lagged.

around 2 weeks ago something happened (no idea what exactly) and now if I don't access these mounts for a small duration (few minutes even) and then try to access them (even with a simple ls or find command) - they hang, and return "Host is down" message. Then if try again and again 1 or 2-3 more times - access is suddenly ok again and everything is fast again.

running "find" under strace for one of these mounts I noticed this is the system call that is responsible for these messages:

19895 newfstatat(AT_FDCWD, "/home/user/work/mountpoint/cifs1/folder/*", 0x55ab02f81b98, AT_SYMLINK_NOFOLLOW) = -1 EHOSTDOWN (Host is down)

I tried running autofs/automount in debug mode w/foreground output (stdout/stderr) but did not notice anything unusual when the issue occurred.

Other people in my workplace utilize the same setup and have no issues at all (most of them access their mounts far less frequently than myself)

I would appreciate any hints or additional things i can check to resolve this!