ipfs-cluster/pstoremgr/pstoremgr.go

232 lines
5.9 KiB
Go
Raw Normal View History

Feat: emancipate Consensus from the Cluster component This commit promotes the Consensus component (and Raft) to become a fully independent thing like other components, passed to NewCluster during initialization. Cluster (main component) no longer creates the consensus layer internally. This has triggered a number of breaking changes that I will explain below. Motivation: Future work will require the possibility of running Cluster with a consensus layer that is not Raft. The "consensus" layer is in charge of maintaining two things: * The current cluster peerset, as required by the implementation * The current cluster pinset (shared state) While the pinset maintenance has always been in the consensus layer, the peerset maintenance was handled by the main component (starting by the "peers" key in the configuration) AND the Raft component (internally) and this generated lots of confusion: if the user edited the peers in the configuration they would be greeted with an error. The bootstrap process (adding a peer to an existing cluster) and configuration key also complicated many things, since the main component did it, but only when the consensus was initialized and in single peer mode. In all this we also mixed the peerstore (list of peer addresses in the libp2p host) with the peerset, when they need not to be linked. By initializing the consensus layer before calling NewCluster, all the difficulties in maintaining the current implementation in the same way have come to light. Thus, the following changes have been introduced: * Remove "peers" and "bootstrap" keys from the configuration: we no longer edit or save the configuration files. This was a very bad practice, requiring write permissions by the process to the file containing the private key and additionally made things like Puppet deployments of cluster difficult as configuration would mutate from its initial version. Needless to say all the maintenance associated to making sure peers and bootstrap had correct values when peers are bootstrapped or removed. A loud and detailed error message has been added when staring cluster with an old config, along with instructions on how to move forward. * Introduce a PeerstoreFile ("peerstore") which stores peer addresses: in ipfs, the peerstore is not persisted because it can be re-built from the network bootstrappers and the DHT. Cluster should probably also allow discoverability of peers addresses (when not bootstrapping, as in that case we have it), but in the meantime, we will read and persist the peerstore addresses for cluster peers in this file, different from the configuration. Note that dns multiaddresses are now fully supported and no IPs are saved when we have DNS multiaddresses for a peer. * The former "peer_manager" code is now a pstoremgr module, providing utilities to parse, add, list and generally maintain the libp2p host peerstore, including operations on the PeerstoreFile. This "pstoremgr" can now also be extended to perform address autodiscovery and other things indepedently from Cluster. * Create and initialize Raft outside of the main Cluster component: since we can now launch Raft independently from Cluster, we have more degrees of freedom. A new "staging" option when creating the object allows a raft peer to be launched in Staging mode, waiting to be added to a running consensus, and thus, not electing itself as leader or doing anything like we were doing before. This additionally allows us to track when the peer has become a Voter, which only happens when it's caught up with the state, something that was wonky previously. * The raft configuration now includes an InitPeerset key, which allows to provide a peerset for new peers and which is ignored when staging==true. The whole Raft initialization code is way cleaner and stronger now. * Cluster peer bootsrapping is now an ipfs-cluster-service feature. The --bootstrap flag works as before (additionally allowing comma-separated-list of entries). What bootstrap does, is to initialize Raft with staging == true, and then call Join in the main cluster component. Only when the Raft peer transitions to Voter, consensus becomes ready, and cluster becomes Ready. This is cleaner, works better and is less complex than before (supporting both flags and config values). We also backup and clean the state whenever we are boostrapping, automatically * ipfs-cluster-service no longer runs the daemon. Starting cluster needs now "ipfs-cluster-service daemon". The daemon specific flags (bootstrap, alloc) are now flags for the daemon subcommand. Here we mimic ipfs ("ipfs" does not start the daemon but print help) and pave the path for merging both service and ctl in the future. While this brings some breaking changes, it significantly reduces the complexity of the configuration, the code and most importantly, the documentation. It should be easier now to explain the user what is the right way to launch a cluster peer, and more difficult to make mistakes. As a side effect, the PR also: * Fixes #381 - peers with dynamic addresses * Fixes #371 - peers should be Raft configuration option * Fixes #378 - waitForUpdates may return before state fully synced * Fixes #235 - config option shadowing (no cfg saves, no need to shadow) License: MIT Signed-off-by: Hector Sanjuan <code@hector.link>
2018-04-28 22:22:23 +00:00
// Package pstoremgr provides a Manager that simplifies handling
// addition, listing and removal of cluster peer multiaddresses from
// the libp2p Host. This includes resolving DNS addresses, decapsulating
// and encapsulating the /p2p/ (/ipfs/) protocol as needed, listing, saving
// and loading addresses.
package pstoremgr
import (
"bufio"
"context"
"fmt"
"os"
"sync"
"time"
"github.com/ipfs/ipfs-cluster/api"
logging "github.com/ipfs/go-log"
host "github.com/libp2p/go-libp2p-host"
peer "github.com/libp2p/go-libp2p-peer"
peerstore "github.com/libp2p/go-libp2p-peerstore"
ma "github.com/multiformats/go-multiaddr"
madns "github.com/multiformats/go-multiaddr-dns"
)
var logger = logging.Logger("pstoremgr")
// Timeouts for network operations triggered by the Manager
var (
DNSTimeout = 2 * time.Second
ConnectTimeout = 10 * time.Second
)
// Manager provides utilities for handling cluster peer addresses
// and storing them in a libp2p Host peerstore.
type Manager struct {
ctx context.Context
host host.Host
peerstoreLock sync.Mutex
peerstorePath string
}
// New creates a Manager with the given libp2p Host and peerstorePath.
// The path indicates the place to persist and read peer addresses from.
// If empty, these operations (LoadPeerstore, SavePeerstore) will no-op.
func New(h host.Host, peerstorePath string) *Manager {
return &Manager{
ctx: context.Background(),
host: h,
peerstorePath: peerstorePath,
}
}
// ImportPeer adds a new peer address to the host's peerstore, optionally
// dialing to it. It will resolve any DNS multiaddresses before adding them.
// The address is expected to include the /ipfs/<peerID> protocol part.
func (pm *Manager) ImportPeer(addr ma.Multiaddr, connect bool) error {
if pm.host == nil {
return nil
}
logger.Debugf("adding peer address %s", addr)
pid, decapAddr, err := api.Libp2pMultiaddrSplit(addr)
if err != nil {
return err
}
pm.host.Peerstore().AddAddr(pid, decapAddr, peerstore.PermanentAddrTTL)
// dns multiaddresses need to be resolved because libp2p only does that
// on explicit bhost.Connect().
if madns.Matches(addr) {
ctx, cancel := context.WithTimeout(pm.ctx, DNSTimeout)
defer cancel()
resolvedAddrs, err := madns.Resolve(ctx, addr)
if err != nil {
logger.Error(err)
return err
}
pm.ImportPeers(resolvedAddrs, connect)
}
if connect {
ctx, cancel := context.WithTimeout(pm.ctx, ConnectTimeout)
defer cancel()
pm.host.Network().DialPeer(ctx, pid)
}
return nil
}
// RmPeer clear all addresses for a given peer ID from the host's peerstore.
func (pm *Manager) RmPeer(pid peer.ID) error {
if pm.host == nil {
return nil
}
logger.Debugf("forgetting peer %s", pid.Pretty())
pm.host.Peerstore().ClearAddrs(pid)
return nil
}
// if the peer has dns addresses, return only those, otherwise
// return all. In all cases, encapsulate the peer ID.
func (pm *Manager) filteredPeerAddrs(p peer.ID) []ma.Multiaddr {
all := pm.host.Peerstore().Addrs(p)
peerAddrs := []ma.Multiaddr{}
peerDNSAddrs := []ma.Multiaddr{}
peerPart, _ := ma.NewMultiaddr(fmt.Sprintf("/ipfs/%s", peer.IDB58Encode(p)))
for _, a := range all {
encAddr := a.Encapsulate(peerPart)
if madns.Matches(encAddr) {
peerDNSAddrs = append(peerDNSAddrs, encAddr)
} else {
peerAddrs = append(peerAddrs, encAddr)
}
}
if len(peerDNSAddrs) > 0 {
return peerDNSAddrs
}
return peerAddrs
}
// PeersAddresses returns the list of multiaddresses (encapsulating the
// /ipfs/<peerID> part) for the given set of peers. For peers for which
// we know DNS multiaddresses, we only return those. Otherwise, we return
// all the multiaddresses known for that peer.
func (pm *Manager) PeersAddresses(peers []peer.ID) []ma.Multiaddr {
if pm.host == nil {
return nil
}
if peers == nil {
return nil
}
var addrs []ma.Multiaddr
for _, p := range peers {
if p == pm.host.ID() {
continue
}
addrs = append(addrs, pm.filteredPeerAddrs(p)...)
}
return addrs
}
// ImportPeers calls ImportPeer for every address in the given slice, using the
// given connect parameter.
func (pm *Manager) ImportPeers(addrs []ma.Multiaddr, connect bool) error {
for _, a := range addrs {
pm.ImportPeer(a, connect)
}
return nil
}
// ImportPeersFromPeerstore reads the peerstore file and calls ImportPeers with
// the addresses obtained from it.
func (pm *Manager) ImportPeersFromPeerstore(connect bool) error {
return pm.ImportPeers(pm.LoadPeerstore(), connect)
}
// LoadPeerstore parses the peerstore file and returns the list
// of addresses read from it.
func (pm *Manager) LoadPeerstore() (addrs []ma.Multiaddr) {
if pm.peerstorePath == "" {
return
}
pm.peerstoreLock.Lock()
defer pm.peerstoreLock.Unlock()
f, err := os.Open(pm.peerstorePath)
if err != nil {
return // nothing to load
}
defer f.Close()
scanner := bufio.NewScanner(f)
for scanner.Scan() {
addrStr := scanner.Text()
if addrStr[0] != '/' {
// skip anything that is not going to be a multiaddress
continue
}
addr, err := ma.NewMultiaddr(addrStr)
if err != nil {
logger.Errorf(
Feat: emancipate Consensus from the Cluster component This commit promotes the Consensus component (and Raft) to become a fully independent thing like other components, passed to NewCluster during initialization. Cluster (main component) no longer creates the consensus layer internally. This has triggered a number of breaking changes that I will explain below. Motivation: Future work will require the possibility of running Cluster with a consensus layer that is not Raft. The "consensus" layer is in charge of maintaining two things: * The current cluster peerset, as required by the implementation * The current cluster pinset (shared state) While the pinset maintenance has always been in the consensus layer, the peerset maintenance was handled by the main component (starting by the "peers" key in the configuration) AND the Raft component (internally) and this generated lots of confusion: if the user edited the peers in the configuration they would be greeted with an error. The bootstrap process (adding a peer to an existing cluster) and configuration key also complicated many things, since the main component did it, but only when the consensus was initialized and in single peer mode. In all this we also mixed the peerstore (list of peer addresses in the libp2p host) with the peerset, when they need not to be linked. By initializing the consensus layer before calling NewCluster, all the difficulties in maintaining the current implementation in the same way have come to light. Thus, the following changes have been introduced: * Remove "peers" and "bootstrap" keys from the configuration: we no longer edit or save the configuration files. This was a very bad practice, requiring write permissions by the process to the file containing the private key and additionally made things like Puppet deployments of cluster difficult as configuration would mutate from its initial version. Needless to say all the maintenance associated to making sure peers and bootstrap had correct values when peers are bootstrapped or removed. A loud and detailed error message has been added when staring cluster with an old config, along with instructions on how to move forward. * Introduce a PeerstoreFile ("peerstore") which stores peer addresses: in ipfs, the peerstore is not persisted because it can be re-built from the network bootstrappers and the DHT. Cluster should probably also allow discoverability of peers addresses (when not bootstrapping, as in that case we have it), but in the meantime, we will read and persist the peerstore addresses for cluster peers in this file, different from the configuration. Note that dns multiaddresses are now fully supported and no IPs are saved when we have DNS multiaddresses for a peer. * The former "peer_manager" code is now a pstoremgr module, providing utilities to parse, add, list and generally maintain the libp2p host peerstore, including operations on the PeerstoreFile. This "pstoremgr" can now also be extended to perform address autodiscovery and other things indepedently from Cluster. * Create and initialize Raft outside of the main Cluster component: since we can now launch Raft independently from Cluster, we have more degrees of freedom. A new "staging" option when creating the object allows a raft peer to be launched in Staging mode, waiting to be added to a running consensus, and thus, not electing itself as leader or doing anything like we were doing before. This additionally allows us to track when the peer has become a Voter, which only happens when it's caught up with the state, something that was wonky previously. * The raft configuration now includes an InitPeerset key, which allows to provide a peerset for new peers and which is ignored when staging==true. The whole Raft initialization code is way cleaner and stronger now. * Cluster peer bootsrapping is now an ipfs-cluster-service feature. The --bootstrap flag works as before (additionally allowing comma-separated-list of entries). What bootstrap does, is to initialize Raft with staging == true, and then call Join in the main cluster component. Only when the Raft peer transitions to Voter, consensus becomes ready, and cluster becomes Ready. This is cleaner, works better and is less complex than before (supporting both flags and config values). We also backup and clean the state whenever we are boostrapping, automatically * ipfs-cluster-service no longer runs the daemon. Starting cluster needs now "ipfs-cluster-service daemon". The daemon specific flags (bootstrap, alloc) are now flags for the daemon subcommand. Here we mimic ipfs ("ipfs" does not start the daemon but print help) and pave the path for merging both service and ctl in the future. While this brings some breaking changes, it significantly reduces the complexity of the configuration, the code and most importantly, the documentation. It should be easier now to explain the user what is the right way to launch a cluster peer, and more difficult to make mistakes. As a side effect, the PR also: * Fixes #381 - peers with dynamic addresses * Fixes #371 - peers should be Raft configuration option * Fixes #378 - waitForUpdates may return before state fully synced * Fixes #235 - config option shadowing (no cfg saves, no need to shadow) License: MIT Signed-off-by: Hector Sanjuan <code@hector.link>
2018-04-28 22:22:23 +00:00
"error parsing multiaddress from %s: %s",
pm.peerstorePath,
err,
)
}
addrs = append(addrs, addr)
}
if err := scanner.Err(); err != nil {
logger.Errorf("reading %s: %s", pm.peerstorePath, err)
}
return addrs
}
// SavePeerstore stores a slice of multiaddresses in the peerstore file, one
// per line.
func (pm *Manager) SavePeerstore(addrs []ma.Multiaddr) {
if pm.peerstorePath == "" {
return
}
pm.peerstoreLock.Lock()
defer pm.peerstoreLock.Unlock()
f, err := os.Create(pm.peerstorePath)
if err != nil {
logger.Errorf(
"could not save peer addresses to %s: %s",
pm.peerstorePath,
err,
)
return
}
defer f.Close()
for _, a := range addrs {
f.Write([]byte(fmt.Sprintf("%s\n", a.String())))
}
}
// SavePeerstoreForPeers calls PeersAddresses and then saves the peerstore
// file using the result.
func (pm *Manager) SavePeerstoreForPeers(peers []peer.ID) {
pm.SavePeerstore(pm.PeersAddresses(peers))
}