diff --git a/README.md b/README.md
index 092ff93..39a60c2 100644
--- a/README.md
+++ b/README.md
@@ -9,4 +9,6 @@ The specifications are defined following the [OpenAPI Initiative](https://www.op
 Main data exchange format is JSON.
 The specifications are defined following the [JSON schema](https://json-schema.org) format standard.
 
-## Database schemas
+## Database schema
+
+Recommended minimal SQL database schema for job meta data.
diff --git a/schema/json/job-metric-data.schema.json b/schema/json/job-metric-data.schema.json
index 8d0cbb5..09f9561 100644
--- a/schema/json/job-metric-data.schema.json
+++ b/schema/json/job-metric-data.schema.json
@@ -21,6 +21,24 @@
             "description": "Measurement interval in seconds",
             "type": "integer"
         },
+        "thresholds": {
+            "description": "Metric thresholds for specific system",
+            "type": "object",
+            "properties": {
+                "peak": {
+                    "type": "number"
+                },
+                "normal": {
+                    "type": "number"
+                },
+                "caution": {
+                    "type": "number"
+                },
+                "alert": {
+                    "type": "number"
+                }
+            }
+        },
         "series": {
             "description": "",
             "type": "array",
diff --git a/schema/sql/jobs-sqlite.sql b/schema/sql/jobs-sqlite.sql
index 60da539..061f14a 100644
--- a/schema/sql/jobs-sqlite.sql
+++ b/schema/sql/jobs-sqlite.sql
@@ -1,9 +1,9 @@
 CREATE TABLE job ( id INTEGER PRIMARY KEY,
- job_id TEXT, user_id TEXT, project_id TEXT, cluster_id TEXT,
- start_time INTEGER, duration INTEGER,
+ job_id TEXT NOT NULL, user_id TEXT NOT NULL, project_id TEXT NOT NULL, cluster_id TEXT NOT NULL,
+ start_time INTEGER NOT NULL, duration INTEGER NOT NULL,
  walltime INTEGER, job_state TEXT,
- num_nodes INTEGER, node_list TEXT, has_profile INTEGER,
- mem_used_max REAL, flops_any_avg REAL, mem_bw_avg REAL, net_bw_avg REAL, file_bw_avg REAL);
+ num_nodes INTEGER NOT NULL, node_list TEXT NOT NULL, has_profile INTEGER NOT NULL,
+ mem_used_max REAL, flops_any_avg REAL, mem_bw_avg REAL, load_avg REAL, net_bw_avg REAL, file_bw_avg REAL);
 CREATE TABLE tag ( id INTEGER PRIMARY KEY, tag_type TEXT, tag_name TEXT);
 CREATE TABLE jobtag ( job_id INTEGER, tag_id INTEGER, PRIMARY KEY (job_id, tag_id),
  FOREIGN KEY (job_id) REFERENCES job (id)  ON DELETE CASCADE ON UPDATE NO ACTION,