diff --git a/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/down.sql b/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/down.sql
new file mode 100644
index 0000000..ec1fdef
--- /dev/null
+++ b/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/down.sql
@@ -0,0 +1,3 @@
+-- This file should undo anything in `up.sql`
+ALTER TABLE jobs DROP CONSTRAINT assigned_worker;
+ALTER TABLE jobs DROP COLUMN assigned_worker_id;
diff --git a/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/up.sql b/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/up.sql
new file mode 100644
index 0000000..92502ed
--- /dev/null
+++ b/server/migrations/2024-03-10-162811_add_assigned_worker_id_to_jobs/up.sql
@@ -0,0 +1,3 @@
+-- Your SQL goes here
+ALTER TABLE jobs ADD COLUMN assigned_worker_id INT;
+ALTER TABLE jobs ADD CONSTRAINT assigned_worker FOREIGN KEY(assigned_worker_id) REFERENCES workers(id);
diff --git a/server/src/models.rs b/server/src/models.rs
index 566d8eb..f6e7570 100644
--- a/server/src/models.rs
+++ b/server/src/models.rs
@@ -50,6 +50,7 @@ pub struct Job {
     pub finish_time: Option<chrono::DateTime<chrono::Utc>>,
     pub error_message: Option<String>,
     pub elapsed_secs: Option<i64>,
+    pub assigned_worker_id: Option<i32>,
 }
 
 #[derive(Insertable)]
diff --git a/server/src/routes.rs b/server/src/routes.rs
index 5c67ab4..ebaf058 100644
--- a/server/src/routes.rs
+++ b/server/src/routes.rs
@@ -1,8 +1,6 @@
-use std::time::Duration;
-
 use crate::{
     api,
-    models::{Job, NewWorker, Pipeline},
+    models::{Job, NewWorker, Pipeline, Worker},
     DbPool,
 };
 use anyhow::Context;
@@ -146,14 +144,25 @@ pub async fn worker_poll(
         use crate::schema::jobs::dsl::*;
         match jobs
             .filter(status.eq("created"))
-            .filter(arch.eq(payload.arch))
+            .filter(arch.eq(&payload.arch))
             .first::<Job>(conn)
             .optional()?
         {
             Some(job) => {
+                // find worker id
+                let worker = crate::schema::workers::dsl::workers
+                    .filter(crate::schema::workers::dsl::hostname.eq(&payload.hostname))
+                    .filter(crate::schema::workers::dsl::arch.eq(&payload.arch))
+                    .first::<Worker>(conn)?;
+
+                // remove if already allocated to the worker
+                diesel::update(jobs.filter(assigned_worker_id.eq(worker.id)))
+                    .set((status.eq("created"), assigned_worker_id.eq(None::<i32>)))
+                    .execute(conn)?;
+
                 // allocate to the worker
                 diesel::update(&job)
-                    .set(status.eq("assigned"))
+                    .set((status.eq("assigned"), assigned_worker_id.eq(worker.id)))
                     .execute(conn)?;
 
                 // get pipeline the job belongs to
@@ -234,6 +243,7 @@ pub async fn worker_job_update(
                     log_url.eq(res.log_url),
                     finish_time.eq(chrono::Utc::now()),
                     elapsed_secs.eq(res.elapsed_secs),
+                    assigned_worker_id.eq(None::<i32>),
                 ))
                 .execute(&mut conn)?;
         }
diff --git a/server/src/schema.rs b/server/src/schema.rs
index 552daef..39a0937 100644
--- a/server/src/schema.rs
+++ b/server/src/schema.rs
@@ -18,6 +18,7 @@ diesel::table! {
         finish_time -> Nullable<Timestamptz>,
         error_message -> Nullable<Text>,
         elapsed_secs -> Nullable<Int8>,
+        assigned_worker_id -> Nullable<Int4>,
     }
 }
 
@@ -48,9 +49,6 @@ diesel::table! {
 }
 
 diesel::joinable!(jobs -> pipelines (pipeline_id));
+diesel::joinable!(jobs -> workers (assigned_worker_id));
 
-diesel::allow_tables_to_appear_in_same_query!(
-    jobs,
-    pipelines,
-    workers,
-);
+diesel::allow_tables_to_appear_in_same_query!(jobs, pipelines, workers,);