| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
| |
Co-authored-by: Simon Rettberg <srett@users.noreply.github.com>
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
Output according messages and abort connection if
the reply didn't contain exactly as many bytes as
requested, instead of going on and running into
desync with the reply stream from the server, which
would just give us confusing messages about header
magic mismatch.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
5ms was apparently way too low and lead to frequent switch-aways from
the local dnbd3-server on heavy load during bootup.
|
| |
|
|
|
| |
Co-authored-by: Jonathan Bauer <jonathan.bauer@rz.uni-freiburg.de>
Co-authored-by: Manuel Messner <mm@skelett.io>
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
Even if we didn't switch because we already use the requested server.
|
| |
|
|
|
|
|
| |
If we switch to a different server when we only have something in
the send list but nothing in the recv list, the send worker would
not have gotten invoked. Now we unconditionally trigger the send
worker when asked to re-queue any pending requests.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Using workqueues frees us from having to manage the lifecycle
of three dedicated threads. Discovery (alt server checks) and
sending keepalive packets is now done using work on the
power efficient system queue. Sending and receiving happens
via dedicated work queues with higher priority.
blk-mq has also been around for quite a while in the kernel,
so switching to it doesn't hurt backwards compatibility.
As the code is now refactored to work more as blk-mq is designed,
backwards compatibility even improved while at the same time
freeing us from an arsenal of macros that were required to make
the blk-mq port look and feel like the old implementation.
For example, the code now compiles on CentOS 7 with kernel 3.10
without requiring special macros to detect the heavily modified
RedHat kernel with all its backported features.
A few other design limitations have been rectified along the way,
e.g. switching to another server now doesn't internally disconnect
from the current one first, which theoretically could lead to a
non-working setup, if the new server isn't reachable and then -
because of some transient network error - switching back also
fails. As the discover-thread was torn down from the disconnect
call, the connection would also not repair itself eventually.
we now establish the new connection in parallel to the old one,
and only if that succeeds do we replace the old one with it,
similar to how the automatic alt-server switch already does it.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
This spams scary red errors to dmesg when really an unreachable alt
server isn't that much of a deal during normal operation. Change the
log level to debug instead.
Might even consider not printing anything at all.
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
When establishing a new connection on a disconnected device, the old
list of alt-servers was retained. This would lead to us connecting to
the wrong server, as the number of newly passed servers was used when
looping over the list of alt-servers to actually connect.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
| |
- Remove the ugly timeout hack that apparently isn't required after all.
- Set a few socket options that appear to make sense in out use case (no
linger, only one SYN retry, NODELAY).
- Adapt socket timeout in panic mode, in case we're on a very bad
connection.
|
| |
|
|
|
|
| |
Remove superflous, reduntant or otherwise useless
information. Use space as separator instead of comma
for better readability and easier parsing in shell etc.
|
| |
|
|
|
|
|
| |
Similar logic already exists in the fuse client:
Count how many times in a row a server was fastest when
measuring RTTs, and lower the switching threshold
more the higher the count gets.
|
| |
|
|
|
|
|
|
| |
Convert dnbd3_host_t to struct sockaddr immediately when
adding alt servers, so we don't have to convert it every time
we establish a connection. Additionally we can now use %pISpc
in printf-like functions instead of having if/else constructs
whenever we want to print an address.
|
| |
|
|
|
| |
This avoids automatically switching back right after adding
and switching to a server.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Formerly, the request that was about to be received was looked up in
the receive queue without removing it, then the request payload was
received from the socket while the lock was not being held, and finally,
the lock was required again and the request removed from the queue.
This is dangrous as another thread can concurrently take the request
from the queue while the receive thread reads the payload from the
socket, leading to a double-free by calling blk_mq_end_request twice.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Previously disconnect was protected against concurrent calls, but connect
wasn't. It was easy to crash the kernel when calling connect and
disconnect IOCTLs in a tight loop concurrently.
A global lock was introduced to make sure only one caller can change the
connection state at a time. dev->connection_lock needs to be aquired
when calling dnbd3_net_connect or _disconnect.
This atomic_t based locking mechanism should be turned into a mutex in a
next step, relying on mutex_trylock for cases where we don't have the
cmpxchg-schedule() loop.
Along the way it was noticed that the send/receive timeouts don't apply
to kernel_connect, which might have been the case in older 3.x kernel
versions. A crude workaround using nonblocking connect has been
introduced to emulate this, but a clean solution for this is welcomed.
Also, devices are now properly closed on module unload.
|
| |
|
|
|
|
|
| |
This patch replaces the __VA_OPT__ varadic macro for the C preprocessor
with a builtin GCC extension for optional varadic arguments. Based on
this change, the dnbd3 Linux kernel module can be also built with GCC
7.5 on a fresh Ubuntu 18.04 with Linux kernel version 4.19.x.
|
| |
|
|
|
|
|
| |
With this change, the dnbd3-client is able to display the selected
server (choosen by the Linux kernel module) after successful probing.
Therefore, the ioctl call OPEN returns the selected server host as
index of the specified dnbd3-server hosts after successful probing.
|
