tell the shardmaster that a group is joining.

do a complete agreement. it might choose the wrong leader initially, and have to re-submit after giving up. entirely gives up after about 10 seconds.

enable/disable a ClientEnd.

attach server i to the net.

detach server i from the net.

send an RPC, wait for the reply. the return value indicates success; false means that no reply was received from the server.

tell the shardmaster that a group is leaving.

connect a ClientEnd to a server. a ClientEnd can only be connected once in its lifetime.

create a client end-point. start the thread that listens and delivers.

check that there's exactly one leader. try a few times in case re-elections are needed.

start a Test. print the Test message. e.g. cfg.begin("Test (2B): RPC counts aren't too high")

the service using Raft (e.g. a k/v server) wants to start agreement on the next command to be appended to Raft's log. if this server isn't the leader

end a Test -- the fact that we got here means there was no failure. print the Passed message, and some performance numbers.

Basic test is as follows: one or more clients submitting Append/Get operations to set of servers for some period of time. After the period is over, t

start or re-start a Raft. if one already exists, "kill" it first. allocate new outgoing port file names, and a new state persister, to isolate previo

Wait blocks until the currently scheduled work has completed. This happens when all tasks have scheduled and completed, the final output have been com

debug() will only print if debugEnabled is true

Create a clerk with clerk specific server names. Give it connections to all of the servers, but for now enable only connections to servers in to[].

return currentTerm and whether this server believes it is the leader.

start i'th server in gi'th group

Maximum log size across all servers

If restart servers, first call ShutdownServer

start a Test. print the Test message. e.g. cfg.begin("Test (2B): RPC counts aren't too high")

should be called with a lock

end a Test -- the fact that we got here means there was no failure. print the Passed message, and some performance numbers.

Shutdown i'th server of gi'th group, by isolating it

Cook up a unique-ish UNIX-domain socket name in /var/tmp. can't use current directory since AFS doesn't support UNIX-domain sockets.

get/put/putappend that keep counts

Register is an RPC method that is called by workers after they have started up to report that they are ready to receive tasks.

RunWorker sets up a connection with the master, registers its address, and waits for tasks to be scheduled.

Checks input file agaist output file: each input number should show up in the output file in string sorted order

how many servers think a log entry is committed?

Workers report back how many RPCs they have processed in the Shutdown reply. Check that they processed at least 1 DoTask RPC.

shut down a Raft server but save its persistent state.

master server name for labrpc.

Sets up 2 partitions with connectivity between servers in each partition.

call() sends an RPC to the rpcname handler on server srv with arguments args, waits for the reply, and leaves the reply in reply. the reply argument s

attach server i to servers listed in to caller must hold cfg.mu

attach server i to servers listed in to caller must hold cfg.mu

detach server i from the servers listed in from caller must hold cfg.mu

detach server i from the servers listed in from caller must hold cfg.mu

Create a clerk with clerk specific server names. Give it connections to all of the servers, but for now enable only connections to servers in to[].

mergeName constructs the name of the output file of reduce task <reduceTask>

removeFile is a simple wrapper around os.Remove that logs errors.

shard server name for labrpc. i'th server of group gid.

wait for at least n servers to commit. but don't wait forever.

the tester calls Kill() when a Raft instance won't be needed again. you are not required to do anything in Kill(), but it might be convenient to (for

should be called with lock

check that no server's log is too big.

save Raft's persistent state to stable storage, where it can later be retrieved after a crash and restart. see paper's Figure 2 for a description of