[[resending with correct cc: - "vfs.kernel.org" just isn't right!]]
xprt->reestablish_timeout is used to cause TCP connection attempts to
back off if the connection fails so as not to hammer the network,
but to still allow immediate connections when there is no reason to
believe there is a problem.
It is not used for the first connection (when transport->sock is NULL)
but only on reconnects.
It is currently set:
a/ to 0 when xs_tcp_state_change finds a state of TCP_FIN_WAIT1
on the assumption that the client has closed the connection
so the reconnect should be immediate when needed.
b/ to at least XS_TCP_INIT_REEST_TO when xs_tcp_state_change
detects TCP_CLOSING or TCP_CLOSE_WAIT on the assumption that the
server closed the connection so a small delay at least is
required.
c/ as above when xs_tcp_state_change detects TCP_SYN_SENT, so that
it is never 0 while a connection has been attempted, else
the doubling will produce 0 and there will be no backoff.
d/ to double is value (up to a limit) when delaying a connection,
thus providing exponential backoff and
e/ to XS_TCP_INIT_REEST_TO in xs_setup_tcp as simple initialisation.
So you can see it is highly dependant on xs_tcp_state_change being
called as expected. However experimental evidence shows that
xs_tcp_state_change does not see all state changes.
("rpcdebug -m rpc trans" can help show what actually happens).
Results show:
TCP_ESTABLISHED is reported when a connection is made. TCP_SYN_SENT
is never reported, so rule 'c' above is never effective.
When the server closes the connection, TCP_CLOSE_WAIT and
TCP_LAST_ACK *might* be reported, and TCP_CLOSE is always
reported. This rule 'b' above will sometimes be effective, but
not reliably.
When the client closes the connection, it used to result in
TCP_FIN_WAIT1, TCP_FIN_WAIT2, TCP_CLOSE. However since commit
f75e674 (SUNRPC: Fix the problem of EADDRNOTAVAIL syslog floods on
reconnect) we don't see *any* events on client-close. I think this
is because xs_restore_old_callbacks is called to disconnect
xs_tcp_state_change before the socket is closed.
In any case, rule 'a' no longer applies.
So all that is left are rule d, which successfully doubles the
timeout which is never rest, and rule e which initialises the timeout.
Even if the rules worked as expected, there would be a problem because
a successful connection does not reset the timeout, so a sequence
of events where the server closes the connection (e.g. during failover
testing) will cause longer and longer timeouts with no good reason.
This patch:
- sets reestablish_timeout to 0 in xs_close thus effecting rule 'a'
- sets it to 0 in xs_tcp_data_ready to ensure that a successful
connection resets the timeout
- sets it to at least XS_TCP_INIT_REEST_TO after it is doubled,
thus effecting rule c
I have not reimplemented rule b and the new version of rule c
seems sufficient.
I suspect other code in xs_tcp_data_ready needs to be revised as well.
For example I don't think connect_cookie is being incremented as often
as it should be.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
[sunrpc: change idle timeout value for the backchannel]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Acked-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
When the call direction is a reply, copy the xid and call direction into the
req->rq_private_buf.head[0].iov_base otherwise rpc_verify_header returns
rpc_garbage.
Signed-off-by: Rahul Iyer <iyer@netapp.com>
Signed-off-by: Mike Sager <sager@netapp.com>
Signed-off-by: Marc Eshel <eshel@almaden.ibm.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[get rid of CONFIG_NFSD_V4_1]
[sunrpc: refactoring of svc_tcp_recvfrom]
[nfsd41: sunrpc: create common send routine for the fore and the back channels]
[nfsd41: sunrpc: Use free_page() to free server backchannel pages]
[nfsd41: sunrpc: Document server backchannel locking]
[nfsd41: sunrpc: remove bc_connect_worker()]
[nfsd41: sunrpc: Define xprt_server_backchannel()[
[nfsd41: sunrpc: remove bc_close and bc_init_auto_disconnect dummy functions]
[nfsd41: sunrpc: eliminate unneeded switch statement in xs_setup_tcp()]
[nfsd41: sunrpc: Don't auto close the server backchannel connection]
[nfsd41: sunrpc: Remove unused functions]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfsd41: change bc_sock to bc_xprt]
[nfsd41: sunrpc: move struct rpc_buffer def into a common header file]
[nfsd41: sunrpc: use rpc_sleep in bc_send_request so not to block on mutex]
[removed cosmetic changes]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[sunrpc: add new xprt class for nfsv4.1 backchannel]
[sunrpc: v2.1 change handling of auto_close and init_auto_disconnect operations for the nfsv4.1 backchannel]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
[reverted more cosmetic leftovers]
[got rid of xprt_server_backchannel]
[separated "nfsd41: sunrpc: add new xprt class for nfsv4.1 backchannel"]
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Cc: Trond Myklebust <trond.myklebust@netapp.com>
[sunrpc: change idle timeout value for the backchannel]
Signed-off-by: Alexandros Batsakis <batsakis@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Acked-by: Trond Myklebust <trond.myklebust@netapp.com>
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
After a bind completes, update the transport instance's address
strings so debugging messages display the current port the transport
is connected to.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
At some point, I recall that rpc_pipe_fs used RPC_DISPLAY_ALL.
Currently there are no uses of RPC_DISPLAY_ALL outside the transport
modules themselves, so we can safely get rid of it.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: Give the "addr" and "port" field less ambiguous names.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up: In addition to using the new generic rpc_ntop() and
rpc_get_port() functions, have the RPC client compute the presentation
address buffer sizes dynamically using kstrdup().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
RPC universal address generation is currently done in several places:
rpcb_clnt.c, nfs4proc.c xprtsock.c, and xprtrdma.c. Remove the
redundant cases that convert a socket address to a universal
address. The nfs4proc.c case takes a pre-formatted presentation
address string, not a socket address, so we'll leave that one.
Because the new uaddr constructor uses the recently introduced
rpc_ntop(), it now supports proper "::" shorthanding for IPv6
addresses. This allows the kernel to register properly formed
universal addresses with the local rpcbind service, in _all_ cases.
The kernel can now also send properly formed universal addresses in
RPCB_GETADDR requests, and support link-local properly when
encoding and decoding IPv6 addresses.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Parameters like the minimum reserved port, and the number of slot entries
should really be module parameters rather than sysctls.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Executes the backchannel task on the RPC state machine using
the existing open connection previously established by the client.
Signed-off-by: Ricardo Labiaga <ricardo.labiaga@netapp.com>
nfs41: Add bc_svc.o to sunrpc Makefile.
[nfs41: bc_send() does not need to be exported outside RPC module]
[nfs41: xprt_free_bc_request() need not be exported outside RPC module]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[Update copyright]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
In the case of -EADDRNOTAVAIL and/or unhandled connection errors, we want
to get rid of the existing socket and retry immediately, just as the
comment says. Currently we end up sleeping for a minute, due to the missing
"break" statement.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Handles RPC replies and backchannel callbacks. Traditionally the NFS
client has expected only RPC replies on its open connections. With
NFSv4.1, callbacks can arrive over an existing open connection.
This patch refactors the old xs_tcp_read_request() into an RPC reply handler:
xs_tcp_read_reply(), a new backchannel callback handler: xs_tcp_read_callback(),
and a common routine to read the data off the transport: xs_tcp_read_common().
The new xs_tcp_read_callback() queues callback requests onto a queue where
the callback service (a separate thread) is listening for the processing.
This patch incorporates work and suggestions from Rahul Iyer (iyer@netapp.com)
and Benny Halevy (bhalevy@panasas.com).
xs_tcp_read_callback() drops the connection when the number of expected
callbacks is exceeded. Use xprt_force_disconnect(), ensuring tasks on
the pending queue are awaken on disconnect.
[nfs41: Keep track of RPC call/reply direction with a flag]
[nfs41: Preallocate rpc_rqst receive buffer for handling callbacks]
Signed-off-by: Ricardo Labiaga <ricardo.labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfs41: sunrpc: xs_tcp_read_callback() should use xprt_force_disconnect()]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[Moves embedded #ifdefs into #ifdef function blocks]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Reading and storing the RPC direction is a three step process.
1. xs_tcp_read_calldir() reads the RPC direction, but it will not store it
in the XDR buffer since the 'struct rpc_rqst' is not yet available.
2. The 'struct rpc_rqst' is obtained during the TCP_RCV_COPY_DATA state.
This state need not necessarily be preceeded by the TCP_RCV_READ_CALLDIR.
For example, we may be reading a continuation packet to a large reply.
Therefore, we can't simply obtain the 'struct rpc_rqst' during the
TCP_RCV_READ_CALLDIR state and assume it's available during TCP_RCV_COPY_DATA.
This patch adds a new TCP_RCV_READ_CALLDIR flag to indicate the need to
read the RPC direction. It then uses TCP_RCV_COPY_CALLDIR to indicate the
RPC direction needs to be saved after the 'struct rpc_rqst' has been allocated.
3. The 'struct rpc_rqst' is obtained by the xs_tcp_read_data() helper
functions. xs_tcp_read_common() then saves the RPC direction in the XDR
buffer if TCP_RCV_COPY_CALLDIR is set. This will happen when we're reading
the data immediately after the direction was read. xs_tcp_read_common()
then clears this flag.
[was nfs41: Skip past the RPC call direction]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfs41: sunrpc: Add RPC direction back into the XDR buffer]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
[nfs41: sunrpc: Don't skip past the RPC call direction]
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
NFSv4.1 callbacks can arrive over an existing connection. This patch adds
the logic to read the RPC call direction (call or reply). It does this by
updating the state machine to look for the call direction invoking
xs_tcp_read_calldir(...) after reading the XID.
[nfs41: Keep track of RPC call/reply direction with a flag]
As per 11/14/08 review of RFC 53/85.
Add a new flag to track whether the incoming message is an RPC call or an
RPC reply. TCP_RPC_REPLY is set in the 'struct sock_xprt' tcp_flags in
xs_tcp_read_calldir() if the message is an RPC reply sent on the forechannel.
It is cleared if the message is an RPC request sent on the back channel.
Signed-off-by: Ricardo Labiaga <Ricardo.Labiaga@netapp.com>
Signed-off-by: Benny Halevy <bhalevy@panasas.com>
Define three accessors to get/set dst attached to a skb
struct dst_entry *skb_dst(const struct sk_buff *skb)
void skb_dst_set(struct sk_buff *skb, struct dst_entry *dst)
void skb_dst_drop(struct sk_buff *skb)
This one should replace occurrences of :
dst_release(skb->dst)
skb->dst = NULL;
Delete skb->dst field
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
See http://bugzilla.kernel.org/show_bug.cgi?id=13034
If the port gets into a TIME_WAIT state, then we cannot reconnect without
binding to a new port.
Tested-by: Petr Vandrovec <petr@vandrovec.name>
Tested-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
As long as one task is holding the socket lock, then calls to
xprt_force_disconnect(xprt) will not succeed in shutting down the socket.
In particular, this would mean that a server initiated shutdown will not
succeed until the lock is relinquished.
In order to avoid the deadlock, we should ensure that xs_tcp_send_request()
closes the socket on EPIPE errors too.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
This fixes a regression against FreeBSD servers as reported by Tomas
Kasparek. Apparently when using RPC over a TCP socket, the FreeBSD servers
don't ever react to the client closing the socket, and so commit
e06799f958 (SUNRPC: Use shutdown() instead of
close() when disconnecting a TCP socket) causes the setup to hang forever
whenever the client attempts to close and then reconnect.
We break the deadlock by adding a 'linger2' style timeout to the socket,
after which, the client will abort the connection using a TCP 'RST'.
The default timeout is set to 15 seconds. A subsequent patch will put it
under user control by means of a systctl.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
If xs_nospace() finds that the socket has disconnected, it attempts to
return ENOTCONN, however that value is then squashed by the callers.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Enforce the comment in xs_tcp_connect_worker4/xs_tcp_connect_worker6 that
we should delay, then retry on certain connection errors.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
While we should definitely return socket errors to the task that is
currently trying to send data, there is no need to propagate the same error
to all the other tasks on xprt->pending. Doing so actually slows down
recovery, since it causes more than one tasks to attempt socket recovery.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
If we get an ECONNREFUSED error, we currently go to sleep on the
'xprt->sending' wait queue. The problem is that no timeout is set there,
and there is nothing else that will wake the task up later.
We should deal with ECONNREFUSED in call_status, given that is where we
also deal with -EHOSTDOWN, and friends.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
...so that we can distinguish between when we need to shutdown and when we
don't. Also remove the call to xs_tcp_shutdown() from xs_tcp_connect(),
since xprt_connect() makes the same test.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
To clear out old state, the UDP connect workers unconditionally invoke
xs_close() before proceeding with a new connect. Nowadays this causes
a spurious wake-up of the task waiting for the connect to complete.
This is a little racey, but usually harmless. The waiting task
immediately retries the connect via a call_bind/call_connect sequence,
which usually finds the transport already in the connected state
because the connect worker has finished in the background.
To avoid a spurious wake-up, factor the xs_close() logic that resets
the underlying socket into a helper, and have the UDP connect workers
call that helper instead of xs_close().
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
If the transport isn't bound, then we should just return ENOTCONN, letting
call_connect_status() and/or call_status() deal with retrying. Currently,
we appear to abort all pending tasks with an EIO error.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
We can Oops in both xs_udp_send_request() and xs_tcp_send_request() if the
call to xs_sendpages() returns an error due to the socket not yet being
set up.
Deal with that situation by returning a new error: ENOTSOCK, so that we
know to avoid dereferencing transport->sock.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Using NIPQUAD() with NIPQUAD_FMT, %d.%d.%d.%d or %u.%u.%u.%u
can be replaced with %pI4
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
gcc warns when using the # modifier with the %p format specifier,
so we can't use this to omit the colons when needed, introduces
%pi6 instead.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The define in kernel.h can be done away with at a later time.
Signed-off-by: Harvey Harrison <harvey.harrison@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If the server sends us an RST error while we're in the TCP_ESTABLISHED
state, then that will not result in a state change, and so the RPC client
ends up hanging forever (see
http://bugzilla.kernel.org/show_bug.cgi?id=11154)
We can intercept the reset by setting up an sk->sk_error_report callback,
which will then allow us to initiate a proper shutdown and retry...
We also make sure that if the send request receives an ECONNRESET, then we
shutdown too...
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Clean up the various different email addresses of mine listed in the code
to a single current and valid address. As Dave says his network merges
for 2.6.28 are now done this seems a good point to send them in where
they won't risk disrupting real changes.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The RPC client uses the rq_xtime field in each RPC request to determine the
round-trip time of the request. Currently, the rq_xtime field is
initialized by each transport just before it starts enqueing a request to
be sent. However, transports do not handle initializing this value
consistently; sometimes they don't initialize it at all.
To make the measurement of request round-trip time consistent for all
RPC client transport capabilities, pull rq_xtime initialization into the
RPC client's generic transport logic. Now all transports will get a
standardized RTT measure automatically, from:
xprt_transmit()
to
xprt_complete_rqst()
This makes round-trip time calculation more accurate for the TCP transport.
The socket ->sendmsg() method can return "-EAGAIN" if the socket's output
buffer is full, so the TCP transport's ->send_request() method may call
the ->sendmsg() method repeatedly until it gets all of the request's bytes
queued in the socket's buffer.
Currently, the TCP transport sets the rq_xtime field every time through
that loop so the final value is the timestamp just before the *last* call
to the underlying socket's ->sendmsg() method. After this patch, the
rq_xtime field contains a timestamp that reflects the time just before the
*first* call to ->sendmsg().
This is consequential under heavy workloads because large requests often
take multiple ->sendmsg() calls to get all the bytes of a request queued.
The TCP transport causes the request to sleep until the remote end of the
socket has received enough bytes to clear space in the socket's local
output buffer. This delay can be quite significant.
The method introduced by this patch is a more accurate measure of RTT
for stream transports, since the server can cause enough back pressure
to delay (ie increase the latency of) requests from the client.
Additionally, this patch corrects the behavior of the RDMA transport, which
entirely neglected to initialize the rq_xtime field. RPC performance
metrics for RDMA transports now display correct RPC request round trip
times.
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Acked-by: Tom Talpey <thomas.talpey@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
NFSv4 requires us to ensure that we break the TCP connection before we're
allowed to retransmit a request. However in the case where we're
retransmitting several requests that have been sent on the same
connection, we need to ensure that we don't interfere with the attempt to
reconnect and/or break the connection again once it has been established.
We therefore introduce a 'connection' cookie that is bumped every time a
connection is broken. This allows requests to track if they need to force a
disconnection.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The condition for exiting from the loop in xs_tcp_send_request() should be
that we find we're not making progress (i.e. number of bytes sent is 0).
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
The rest of the networking layer uses SOCK_ASYNC_NOSPACE to signal whether
or not we have someone waiting for buffer memory. Convert the SUNRPC layer
to use the same idiom.
Remove the unlikely()s in xs_udp_write_space and xs_tcp_write_space. In
fact, the most common case will be that there is nobody waiting for buffer
space.
SOCK_NOSPACE is there to tell the TCP layer whether or not the cwnd was
limited by the application window. Ensure that we follow the same idiom as
the rest of the networking layer here too.
Finally, ensure that we clear SOCK_ASYNC_NOSPACE once we wake up, so that
write_space() doesn't keep waking things up on xprt->pending.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Neil Brown