I have an application that, when it has data to transmit, uses epoll to know whether a given TCP socket can be written to.
What I'm observing is that as the far-end of the TCP connection falls behind, and the send buffer of the TCP socket begins to fill, the frequency with which epoll returns an EPOLLOUT event appears to experience exponential backoff. This behavior happens prior to receiving an EAGAIN from the socket write.
The application is using EPOLLONESHOT, and makes an EPOLL_CTL_MOD call to rearm the EPOLLOUT event after each occurrence. But as I noted above, each subsequent occurrence is exponentially later (I had a progression of 40ms, 80ms, 160ms, 320ms, 640ms, 1280ms, etc), up until EAGAIN finally happens.
Is this an undocumented feature of epoll? Can it be disabled? It's a problem because the data is getting stale and I would prefer to discard it rather than send it late.
Thanks in advance.
No, but TCP does. epoll() blocks for at most the timeout you specify, and not a moment longer.
Related
If a TCP connection is established between two hosts (A & B), and lets say host A has sent 5 octets to host B, and then the host B crashes (due to unknown reason).
The host A will wait for acknowledgments, but on not getting them, will resend octets and also reduce the sender window size.
This will repeat couple times till the window size shrinks to zero because of packet loss. My question is, what will happen next?
In this case, TCP eventually times out waiting for the ack's and return an error to the application. The application have to read/recv from the TCP socket to learn about that error, a subsequent write/send call will fail as well. Up till the point that TCP determined that the connection is gone, write/send calls will not fail, they'll succeed as seen from the application or block if the socket buffer is full.
In the case your host B vanishes after it has sent its ACKs, host A will not learn about that until it sends something to B, which will eventually also time out, or result in an ICMP error. (Typically the first write/send call will not fail as TCP will not fail the connection immediately, and keep in mind that write/send calls does not wait for ACKs until they complete).
Note also that retransmission does not reduce the window size.
Please follow this link
now a very simple answer to your question in my view is, The connection will be timed out and will be closed. another possibility that exists is that some ICMP error might be generated due to due un-responsive machine.
Also, if the crashed machine is online again, then the procedure described in the link i just pasted above will be observed.
Depends on the OS implementation. In short it will wait for ACK and resend packets until it times out. Then your connection will be torn down. To see exactly what happens in Linux look here other OSes follow similar algorithm.
in your case, A FIN will be generated (by the surviving node) and connection will eventually migrate to CLOSED state. If you keep grep-ing for netstat output on the destination ip address, you will watch the migration from ESTABLISHED state to TIMED_WAIT and then finally disappear.
In your case, this will happen since TCP keeps a timer to get the ACK for the packet it has sent. This timer is not long enough so detection will happen pretty quickly.
However, if the machine B dies after A gets ACK and after that A doesn't send anything, then the above timer can't detect the same event, however another timer (calls idle timeout) will detect that condition and connection will close then. This timeout period is high by default. But normally this is not the case, machine A will try to send stuff in between and will detect the error condition in send path.
In short, TCP is smart enough to close the connection by itself (and let application know about it) except for one case (Idle timeout: which by default is very high).
cforfun
In normal cases, each side terminating its end of the connectivity by sending a special message with a FIN(finish) bit set.
The device receiving this FIN responds with an acknowledgement to the FIN to indicate that it has been received.
The connection as a whole is not considered terminated until both the devices complete the shut down procedure by sending an FIN and receiving an acknowledgement.
I am designing and testing a client server program based on TCP sockets(Internet domain). Currently , I am testing it on my local machine and not able to understand the following about SIGPIPE.
*. SIGPIPE appears quite randomly. Can it be deterministic?
The first tests involved single small(25 characters) send operation from client and corresponding receive at server. The same code, on the same machine runs successfully or not(SIGPIPE) totally out of my control. The failure rate is about 45% of times(quite high). So, can I tune the machine in any way to minimize this.
**. The second round of testing was to send 40000 small(25 characters) messages from the client to the server(1MB of total data) and then the server responding with the total size of data it actually received. The client sends data in a tight loop and there is a SINGLE receive call at the server. It works only for a maximum of 1200 bytes of total data sent and again, there are these non deterministic SIGPIPEs, about 70% times now(really bad).
Can some one suggest some improvement in my design(probably it will be at the server). The requirement is that the client shall be able to send over medium to very high amount of data (again about 25 characters each message) after a single socket connection has been made to the server.
I have a feeling that multiple sends against a single receive will always be lossy and very inefficient. Shall we be combining the messages and sending in one send() operation only. Is that the only way to go?
SIGPIPE is sent when you try to write to an unconnected pipe/socket. Installing a handler for the signal will make send() return an error instead.
signal(SIGPIPE, SIG_IGN);
Alternatively, you can disable SIGPIPE for a socket:
int n = 1;
setsockopt(thesocket, SOL_SOCKET, SO_NOSIGPIPE, &n, sizeof(n));
Also, the data amounts you're mentioning are not very high. Likely there's a bug somewhere that causes your connection to close unexpectedly, giving a SIGPIPE.
SIGPIPE is raised because you are attempting to write to a socket that has been closed. This does indicate a probable bug so check your application as to why it is occurring and attempt to fix that first.
Attempting to just mask SIGPIPE is not a good idea because you don't really know where the signal is coming from and you may mask other sources of this error. In multi-threaded environments, signals are a horrible solution.
In the rare cases were you cannot avoid this, you can mask the signal on send. If you set the MSG_NOSIGNAL flag on send()/sendto(), it will prevent SIGPIPE being raised. If you do trigger this error, send() returns -1 and errno will be set to EPIPE. Clean and easy. See man send for details.
From what I know, a blocking receive on a TCP socket does not always detect a connection error (due either to a network failure or to a remote-endpoint failure) by returning a -1 value or raising an IO exception: sometimes it could just hang indefinitely.
One way to manage this problem is to set a timeout for the blocking receive. In case an upper bound for the reception time is known, this bound could be set as timeout and the connection could be considered lost simply when the timeout expires; when such an upper bound is not known a priori, for example in a pub-sub system where a connection stays open to receive publications, the timeout to be set would be somewhat arbitrary but its expiration could trigger a ping/pong request to verify that the connection (and the endpoint too) is still up.
I wonder whether the use of asynchronous receive also manages the problem of detecting a connection failure. In boost::asio I would call socket::asynch_read_some() registering an handler to be asynchronously called, while in java.nio I would configure the channel as non-blocking and register it to a selector with an OP_READ interest flag. I imagine that a correct connection-failure detection would mean that, in the first case the handler would be called with a non-0 error_code, while in the second case the selector would select the faulty channel but a subsequent read() on the channel would either return -1 or throw an IOException.
Is this behaviour guaranteed with asynchronous receive, or could there be scenarios where after a connection failure, for example, in boost::asio the handler will never be called or in java.nio the selector will never select the channel?
Thank you very much.
I believe you're referring to the TCP half-open connection problem (the RFC 793 meaning of the term). Under this scenario, the receiving OS will never receive indication of the lost connection, so it will never notify the app. Whether the app is readding synchronously or asynchronously doesn't enter into it.
The problem occurs when the transmitting side of the connection somehow is no longer aware of the network connection. This can happen, for example, when
the transmitting OS abruptly terminates/restarts (power outage, OS failure/BSOD, etc.).
the transmitting side closes its side while there is a network disruption between the two sides and cleans up its side: e.g transmitting OS reboots cleanly during disruption, transmitting Windows OS is unplugged from the network
When this happens, the receiving side may be waiting for data or a FIN that will never come. Unless the receiving side sends a message, there's no way for it to realize the transmitting side is no longer aware of the receiving side.
Your solution (a timeout) is one way to address the issue, but it should include sending a message to the transmitting side. Again, it doesn't matter the read is synchronous or asynchronous, just that it doesn't read and wait indefinitely for data or a FIN. Another solution is using a TCP KEEPALIVE feature that is supported by some TCP stacks. But the hard part of any generalized solution is usually determining a proper timeout, since the timeout is highly dependent on characteristics of the specific application.
Because of how TCP works, you will typically have to send data in order to notice a hard connection failure, to find out that no ACK packet will ever be returned. Some protocols attempt to identify conditions like this by periodically using a keep-alive or ping packet: if one side does not receive such a packet in X time (and perhaps after trying and failing one itself), it can consider the connection dead.
To answer your question, blocking and non-blocking receive should perform identically except for the act of blocking itself, so both will suffer from this same issue. In order to make sure that you can detect a silent failure from the remote host, you'll have to use a form of keep-alive like I described.
I was wondering how tcp/ip communication is implemented in unix. When you do a send over the socket, does the tcp/level work (assembling packets, crc, etc) get executed in the same execution context as the calling code?
Or, what seems more likely, a message is sent to some other daemon process responsible for tcp communication? This process then takes the message and performs the requested work of copying memory buffers and assembling packets etc.? So, the calling code resumes execution right away and tcp work is done in parallel? Is this correct?
Details would be appreciated. Thanks!
The TCP/IP stack is part of your kernel. What happens is that you call a helper method which prepares a "kernel trap". This is a special kind of exception which puts the CPU into a mode with more privileges ("kernel mode"). Inside of the trap, the kernel examines the parameters of the exception. One of them is the number of the function to call.
When the function is called, it copies the data into a kernel buffer and prepares everything for the data to be processed. Then it returns from the trap, the CPU restores registers and its original mode and execution of your code resumes.
Some kernel thread will pick up the copy of the data and use the network driver to send it out, do all the error handling, etc.
So, yes, after copying the necessary data, your code resumes and the actual data transfer happens in parallel.
Note that this is for TCP packets. The TCP protocol does all the error handling and handshaking for you, so you can give it all the data and it will know what to do. If there is a problem with the connection, you'll notice only after a while since the TCP protocol can handle short network outages by itself. That means you'll have "sent" some data already before you'll get an error. That means you will get the error code for the first packet only after the Nth call to send() or when you try to close the connection (the close() will hang until the receiver has acknowledged all packets).
The UDP protocol doesn't buffer. When the call returns, the packet is on it's way. But it's "fire and forget", so you only know that the driver has put it on the wire. If you want to know whether it has arrived somewhere, you must figure out a way to achieve that yourself. The usual approach is have the receiver send an ack UDP packet back (which also might get lost).
No - there is no parallel execution. It is true that the execution context when you're making a system call is not the same as your usual execution context. When you make a system call, such as for sending a packet over the network, you must switch into the kernel's context - the kernel's own memory map and stack, instead of the virtual memory you get inside your process.
But there are no daemon processes magically dispatching your call. The rest of the execution of your program has to wait for the system call to finish and return whatever values it will return. This is why you can count on return values being available right away when you return from the system call - values like the number of bytes actually read from the socket or written to a file.
I tried to find a nice explanation for how the context switch to kernel space works. Here's a nice in-depth one that even focuses on architecture-specific implementation:
http://www.ibm.com/developerworks/linux/library/l-system-calls/
When using a TCP socket, what does
shutdown(sock, SHUT_RD);
actually do? Does it just make all recv() calls return an error code? If so, which error code?
Does it cause any packets to be sent by the underlying TCP connection? What happens to any data that the other side sends at this point - is it kept, and the window size of the connection keeps shrinking until it gets to 0, or is it just discarded, and the window size doesn't shrink?
Shutting down the read side of a socket will cause any blocked recv (or similar) calls to return 0 (indicating graceful shutdown). I don't know what will happen to data currently traveling up the IP stack. It will most certainly ignore data that is in-flight from the other side. It will not affect writes to that socket at all.
In fact, judicious use of shutdown is a good way to ensure that you clean up as soon as you're done. An HTTP client that doesn't use keepalive can shutdown the write-side as soon as it is done sending the request, and a server that sees Connection: closed can likewise shutdown the read-side as soon as it is done receiving the request. This will cause any further erroneous activity to be immediately obvious, which is very useful when writing protocol-level code.
Looking at the Linux source code, shutdown(sock, SHUT_RD) doesn't seem to cause any state changes to the socket. (Obviously, shutdown(sock, SHUT_WR) causes FIN to be set.)
I can't comment on the window size changes (or lack thereof). But you can write a test program to see. Just make your inetd run a chargen service, and connect to it. :-)
shutdown(,SHUT_RD) does not have any counterpart in TCP protocol, so it is pretty much up to implementation how to behave when someone writes to a connection where other side indicated that it will not read or when you try to read after you declared that you wont.
On slightly lower level it is beneficial to remember that TCP connection is a pair of flows using which peers send data until they declare that they are done (by SHUT_WR which sends FIN). And these two flows are quite independent.
I test shudown(sock,SHUT_RD) on Ubuntu 12.04. I find that when you call shutdown(sock,SHUT_RD) if there are no any type of data(include FIN....) in the TCP buffer, the successive read call will return 0(indicates end of stream). But if there are some data which arrived before or after shutdown function, read call will process normally as if shutdown function was not called. It seems that shutdown(sock,SHUT_RD) doesn't cause any TCP states changed to the socket
It has two effects, one of them platform-dependent.
recv() will return zero, indicating end of stream.
Any further writes to the connection by the peer will either be (a) silently thrown away by the receiver (BSD), (b) be buffered by the receiver and eventually cause send() to block or return -1/EAGAIN/EWOULDBLOCK (Linux), or (c) cause the receiver to send an RST (Windows).
shutdown(sock, SHUT_RD) causes any writer to the socket to receive a sigpipe signal.
Any further reads using the read system call will return a -1 and set errno to EINVAL.
The use of recv will return a -1 and set errno to indicate the error (probably ENOTCONN or ENOTSOCK).