- Manvitha Potluri

Client-Server Communication It can be performed in two ways 1. Client-server communication using TCP 2. Client-server communication using RPC

 A reliable point-to-point communication can be established by using TCP protocol.  TCP masks omission failures.  TCP does not mask crash failures.

RPC- Remote procedure calls  The goal of RPC is to hide communication by making remote procedure calls that look just like local ones.  Its difficult to mask RPC.

There are 5 classes of failures in RPC 1. The client unable to locate server. 2. The request message from client to server is lost. 3. The server crashes after receiving a request. 4. The reply message from the server to the client is lost. 5. The client crashes after sending a request.

 Problem- When server goes down  Solution- error raise an exception Java- division by zero C- signal handlers  Drawback- Not every language has exceptions and signals.  Exception destroys transparency.

 The operating system starts a timer when the stub is generated and sends a request. If response is not received before timer expires, then a new request is sent.  Lost message – works fine on retransmission.  Many requests sent- cannot locate server.  If request is not lost, we should make sure server knows that its a retransmission.

 Some messages can be retransmitted any number of times without any loss.  Some retransmissions cause severe loss.  Solution- client assigns sequence number on requests made by client.  Drawback- Server maintains administration on each client. How long to maintain.

 The server experiences 3 types of situations.

 1. Atleast once – RPC carried out >1 time.  2. Atmost once- RPC carried out<=1 time.  Example – print text request

 There are four strategies  1. Never- client will never issue a request at the risk of text not being printed.  2. Always- It can reissue a request but results in printing twice.  3. When not ack- client decides to reissue request to server when there is no acknowledgement that the request is sent.  3. When ack- request is retransmitted.

 With two strategies for server and four for client, there are more combinations to occur.  These are represented by M-send message completion, P- Print text, C- crash.

 Orphans waste CPU cycles, lock files or tie up valuable resources. Even if the client reboots, the confusion persists.  There are mainly four solutions proposed

 1. Orphan extermination- Before the client stub sends the request, a log is created which can survive crashes. After reboot the log is checked and orphan is explicitly killed. Disadvantage- storing data in logs, locating and killing is a heavy task. Sometimes it may fail and result in grand orphans.  2. Reincarnation- Here time is divided into sequentially timed epochs. When it reboots, a new epoch is started and old orphans are removed.

 3. Gentle reincarnation- Each machine checks to see if there are any remote computations running locally, and tries to find their parent. If not found it terminates.  4. Another solution is that, each RPC is given a certain amount of time. If it cannot finish another quantum is requested. After the crash it waits for certain time to reboot where orphans are terminated.

 We have discussed about TCP and RPC methods. But these don’t provide high failure transparency.  In our future we can deal with ADSV(shared variable abstraction) for fault tolerant communication.