Reason: restore is a fdbserver who does not register with CC.
The new failure monitor changes how connection works for client and server.
For client, it does not connect to CC to get connected.
For server, it has to connect to CC to get connected.
Restore worker becomes the special role that behaves like a client but is a server.
So that we can use it from a piece of flow code without breaking module
boundaries.
Also rename generated-constants to crc32c-generated-constants so that
it's more apparent that they're related files.
If there a peer with non-zero reference, keep trying to reconnect when
a connection fails. We already make sure it doesn't happen only once
every 2 seconds.
This patch removes active failure monitoring at server processes.
Hence like client processes, servers no longer require continuously
publishing their membership to cluster controller.
When a process is marked as failed, we still need to know if it back
up at certain point, particularly when the reference count is
incremented. In that case, loadBalance may see AllAlternativesFailed
as failed. To overcome this problem, whenever peer references is
incremented and and the address is marked as failed, connectionKeeper
will bypass waiting for data and connects immediately to check if the
process is back up.
Fix several such reports from ubsan
E.g.
/Users/anoyes/workspace/foundationdb/flow/Arena.h:794:16: runtime error: null pointer passed as argument 1, which is declared to never be null
If client is incompatible, connectionMonitor relies on peer->resetPing
to be triggered whenever data is received to prevent ping timeout.
The server stopped sending pings since 6.2 which meant resetPing
doesn't get triggered.
Instead try pinging the client and let that decide whether the client
is alive or not. Ideally, it should always be failed since a well
behaved client would have closed the connection.
* This will allow client to continue monitoring peer connections while
connection stays open, so that there is no period of "uncertainity"
without previous no-monitoring approach.
* Use multiplier for incoming connection idle timeout
* Update idle connection timeout values and leaked connection timeout in
simulator.
This patch does two changes to connection monitoring:
1. Connection monitoring at client side will check if the connection
has been stayed idle for some time. If connection is unused for a
while, we close the connection. There is some weirdness involved here
as ping messages are by themselves are connection traffic. We get over
this by making it two-phase process, first being checking idle
reliable traffic, followed by disabling pings and then checking for
idle unreliable traffic.
2. Connection monitoring of clients from server will no longer send
pings to clients. Instead, it keep monitor the received bytes and
close after certain period of inactivity.
RequestStream add another count to peerReference, which means as long
as ConnectionMonitor is alive, we'll never get peerReference=0 keeping
unnecessary connections potentially alive.
The constructor of FlowReceiver which handled reference counting
peerReferences relied on calling a virtual method from constructor
whose behaviour isn't correct. This patch, bubbles down result of that
virtual method from derived constructor to base contructor.
For large packet, allocate sizeof(uint32_t) more bytes for next packet size.
Also add knob MIN_PACKET_BUFFER_FREE_BYTES, which is used to trigger allocation
of a new arena when free bytes are lower than this threshold.
On the sending side, a large packet is split into smaller pieces. On the
receiving side, use packet length to allocate buffer to avoid multiple memcpy
and allocations.
This fixes#1214
The basic idea is that ProtocolVersion is now its own type. This
alone is an improvement as it makes many things more typesafe. For
each version, we can now add breaking features (for example Fearless).
After that, there's no need to test against actual (confusing) version
numbers. Instead a developer can simply test
`protocolVersion->hasFearless()` and this will return true iff the
protocolVersion is newer than the newest version that didn't support
fearless.
This patch removes the need for clients to continuously contact
cluster coordinator for failure monitoring information. Instead, it
uses the FlowTransport to monitor the statuses of peers and update
FailureMonitor accordingly.
This commit includes:
- The flatbuffers implementation
- A draft on how it should be used for network messages
- A serializer that can be used independently
What is missing:
- All root objects will need a file identifier
- Many special classes can not be serialized yet as the
corresponding traits are not yet implemented
- Object serialization can not yet be turned on (this will
need a network option)
Vishesh Yadav