Skip to content

fix: taint logic for job nodes with no nomad node #774

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
May 15, 2024
Merged

fix: taint logic for job nodes with no nomad node #774

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
152 changes: 118 additions & 34 deletions svc/pkg/cluster/worker/src/workers/datacenter_scale.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -182,7 +182,13 @@ async fn inner(
};

scale_servers(&ctx, tx, &mut msgs, &servers, &pool_ctx).await?;
drain_tainted_servers(&ctx, tx, &mut msgs, &servers, &pool_ctx).await?;

match pool_ctx.pool_type {
backend::cluster::PoolType::Job => {
cleanup_tainted_job_servers(&ctx, tx, &mut msgs, &servers, &pool_ctx).await?
}
_ => cleanup_tainted_servers(&ctx, tx, &mut msgs, &servers, &pool_ctx).await?,
}
}

destroy_drained_servers(&ctx, tx, &mut msgs, &servers).await?;
Expand Down Expand Up @@ -398,7 +404,7 @@ async fn scale_up_servers<'a, I: Iterator<Item = &'a Server> + Clone>(
Ok(())
}

async fn drain_tainted_servers(
async fn cleanup_tainted_job_servers(
ctx: &OperationContext<()>,
tx: &mut sqlx::Transaction<'_, sqlx::Postgres>,
msgs: &mut Vec<MsgFuture>,
Expand All @@ -416,43 +422,121 @@ async fn drain_tainted_servers(
.filter(|server| server.is_tainted);
let active_tainted_count = active_tainted_servers_in_pool.clone().count();

// For job servers the "active" servers we count are ones with nomad successfully connected. Otherwise we
// count servers that have successfully installed
let active_untainted_count = match pctx.pool_type {
backend::cluster::PoolType::Job => active_servers_in_pool
.clone()
.filter(|server| server.has_nomad_node)
.filter(|server| !server.is_tainted)
.count(),
_ => active_servers_in_pool
.clone()
.filter(|server| server.is_installed)
.filter(|server| !server.is_tainted)
.count(),
};
// For job servers the "active" servers we count are ones with nomad successfully connected
let active_untainted_count = active_servers_in_pool
.clone()
.filter(|server| server.has_nomad_node)
.filter(|server| !server.is_tainted)
.count();

// Amount of tainted servers that need to be deleted or drained
// tainted - (desired - running) -> tainted + running - desired
let drain_count =
let removal_count =
(active_tainted_count + active_untainted_count).saturating_sub(pctx.desired_count);

tracing::info!(
?pctx.pool_type,
desired_count=%pctx.desired_count,
%active_untainted_count,
%active_tainted_count,
%drain_count,
"draining tainted servers",
);
let (nomad_servers, without_nomad_servers) =
active_tainted_servers_in_pool.partition::<Vec<_>, _>(|server| server.has_nomad_node);

drain_servers(
ctx,
tx,
msgs,
active_tainted_servers_in_pool
.take(drain_count)
.map(|server| server.server_id),
)
.await?;
let destroy_count = removal_count.min(without_nomad_servers.len());
let drain_count = removal_count - destroy_count;

if destroy_count != 0 {
tracing::info!(
pool_type=?pctx.pool_type,
desired_count=%pctx.desired_count,
%active_untainted_count,
%active_tainted_count,
%destroy_count,
"destroying tainted servers",
);

destroy_servers(
ctx,
tx,
msgs,
without_nomad_servers
.iter()
.take(destroy_count)
.map(|server| server.server_id),
)
.await?;
}

if drain_count != 0 {
tracing::info!(
pool_type=?pctx.pool_type,
desired_count=%pctx.desired_count,
%active_untainted_count,
%active_tainted_count,
%drain_count,
"draining tainted servers",
);

drain_servers(
ctx,
tx,
msgs,
nomad_servers
.iter()
.take(drain_count)
.map(|server| server.server_id),
)
.await?;
}

Ok(())
}

async fn cleanup_tainted_servers(
ctx: &OperationContext<()>,
tx: &mut sqlx::Transaction<'_, sqlx::Postgres>,
msgs: &mut Vec<MsgFuture>,
servers: &[Server],
pctx: &PoolCtx,
) -> GlobalResult<()> {
// Includes tainted and normal servers
let active_servers_in_pool = servers
.iter()
.filter(|server| server.pool_type == pctx.pool_type)
.filter(|server| matches!(server.drain_state, DrainState::None));

let active_tainted_servers_in_pool = active_servers_in_pool
.clone()
.filter(|server| server.is_tainted);
let active_tainted_count = active_tainted_servers_in_pool.clone().count();

// Count servers that have successfully installed
let active_untainted_count = active_servers_in_pool
.clone()
.filter(|server| server.is_installed)
.filter(|server| !server.is_tainted)
.count();

// Amount of tainted servers that need to be drained
// tainted - (desired - running) -> tainted + running - desired
let drain_count =
(active_tainted_count + active_untainted_count).saturating_sub(pctx.desired_count);

if drain_count != 0 {
tracing::info!(
pool_type=?pctx.pool_type,
desired_count=%pctx.desired_count,
%active_untainted_count,
%active_tainted_count,
%drain_count,
"draining tainted servers",
);

drain_servers(
ctx,
tx,
msgs,
active_tainted_servers_in_pool
.take(drain_count)
.map(|server| server.server_id),
)
.await?;
}

Ok(())
}
Expand Down