Allow to change compute safekeeper list without restart.

- Add --safekeepers option to neon_local reconfigure
- Add it to python Endpoint reconfigure
- Implement config reload in walproposer by restarting the whole bgw when
  safekeeper list changes.

ref https://github.com/neondatabase/neon/issues/6341
This commit is contained in:
Arseny Sher
2024-06-17 16:23:07 +03:00
committed by Arseny Sher
parent d557002675
commit 6f20a18e8e
6 changed files with 139 additions and 55 deletions

View File

@@ -848,20 +848,13 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
let allow_multiple = sub_args.get_flag("allow-multiple");
// If --safekeepers argument is given, use only the listed safekeeper nodes.
let safekeepers =
if let Some(safekeepers_str) = sub_args.get_one::<String>("safekeepers") {
let mut safekeepers: Vec<NodeId> = Vec::new();
for sk_id in safekeepers_str.split(',').map(str::trim) {
let sk_id = NodeId(u64::from_str(sk_id).map_err(|_| {
anyhow!("invalid node ID \"{sk_id}\" in --safekeepers list")
})?);
safekeepers.push(sk_id);
}
safekeepers
} else {
env.safekeepers.iter().map(|sk| sk.id).collect()
};
// If --safekeepers argument is given, use only the listed
// safekeeper nodes; otherwise all from the env.
let safekeepers = if let Some(safekeepers) = parse_safekeepers(sub_args)? {
safekeepers
} else {
env.safekeepers.iter().map(|sk| sk.id).collect()
};
let endpoint = cplane
.endpoints
@@ -965,7 +958,10 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
})
.collect::<Vec<_>>()
};
endpoint.reconfigure(pageservers, None).await?;
// If --safekeepers argument is given, use only the listed
// safekeeper nodes; otherwise all from the env.
let safekeepers = parse_safekeepers(sub_args)?;
endpoint.reconfigure(pageservers, None, safekeepers).await?;
}
"stop" => {
let endpoint_id = sub_args
@@ -987,6 +983,23 @@ async fn handle_endpoint(ep_match: &ArgMatches, env: &local_env::LocalEnv) -> Re
Ok(())
}
/// Parse --safekeepers as list of safekeeper ids.
fn parse_safekeepers(sub_args: &ArgMatches) -> Result<Option<Vec<NodeId>>> {
if let Some(safekeepers_str) = sub_args.get_one::<String>("safekeepers") {
let mut safekeepers: Vec<NodeId> = Vec::new();
for sk_id in safekeepers_str.split(',').map(str::trim) {
let sk_id = NodeId(
u64::from_str(sk_id)
.map_err(|_| anyhow!("invalid node ID \"{sk_id}\" in --safekeepers list"))?,
);
safekeepers.push(sk_id);
}
Ok(Some(safekeepers))
} else {
Ok(None)
}
}
fn handle_mappings(sub_match: &ArgMatches, env: &mut local_env::LocalEnv) -> Result<()> {
let (sub_name, sub_args) = match sub_match.subcommand() {
Some(ep_subcommand_data) => ep_subcommand_data,
@@ -1590,7 +1603,7 @@ fn cli() -> Command {
.about("Start postgres.\n If the endpoint doesn't exist yet, it is created.")
.arg(endpoint_id_arg.clone())
.arg(endpoint_pageserver_id_arg.clone())
.arg(safekeepers_arg)
.arg(safekeepers_arg.clone())
.arg(remote_ext_config_args)
.arg(create_test_user)
.arg(allow_multiple.clone())
@@ -1599,6 +1612,7 @@ fn cli() -> Command {
.subcommand(Command::new("reconfigure")
.about("Reconfigure the endpoint")
.arg(endpoint_pageserver_id_arg)
.arg(safekeepers_arg)
.arg(endpoint_id_arg.clone())
.arg(tenant_id_arg.clone())
)

View File

@@ -499,6 +499,23 @@ impl Endpoint {
.join(",")
}
/// Map safekeepers ids to the actual connection strings.
fn build_safekeepers_connstrs(&self, sk_ids: Vec<NodeId>) -> Result<Vec<String>> {
let mut safekeeper_connstrings = Vec::new();
if self.mode == ComputeMode::Primary {
for sk_id in sk_ids {
let sk = self
.env
.safekeepers
.iter()
.find(|node| node.id == sk_id)
.ok_or_else(|| anyhow!("safekeeper {sk_id} does not exist"))?;
safekeeper_connstrings.push(format!("127.0.0.1:{}", sk.get_compute_port()));
}
}
Ok(safekeeper_connstrings)
}
pub async fn start(
&self,
auth_token: &Option<String>,
@@ -523,18 +540,7 @@ impl Endpoint {
let pageserver_connstring = Self::build_pageserver_connstr(&pageservers);
assert!(!pageserver_connstring.is_empty());
let mut safekeeper_connstrings = Vec::new();
if self.mode == ComputeMode::Primary {
for sk_id in safekeepers {
let sk = self
.env
.safekeepers
.iter()
.find(|node| node.id == sk_id)
.ok_or_else(|| anyhow!("safekeeper {sk_id} does not exist"))?;
safekeeper_connstrings.push(format!("127.0.0.1:{}", sk.get_compute_port()));
}
}
let safekeeper_connstrings = self.build_safekeepers_connstrs(safekeepers)?;
// check for file remote_extensions_spec.json
// if it is present, read it and pass to compute_ctl
@@ -740,6 +746,7 @@ impl Endpoint {
&self,
mut pageservers: Vec<(Host, u16)>,
stripe_size: Option<ShardStripeSize>,
safekeepers: Option<Vec<NodeId>>,
) -> Result<()> {
let mut spec: ComputeSpec = {
let spec_path = self.endpoint_path().join("spec.json");
@@ -774,6 +781,12 @@ impl Endpoint {
spec.shard_stripe_size = stripe_size.map(|s| s.0 as usize);
}
// If safekeepers are not specified, don't change them.
if let Some(safekeepers) = safekeepers {
let safekeeper_connstrings = self.build_safekeepers_connstrs(safekeepers)?;
spec.safekeeper_connstrings = safekeeper_connstrings;
}
let client = reqwest::Client::builder()
.timeout(Duration::from_secs(30))
.build()