diff --git a/utils/crawl-locust.py b/utils/crawl-locust.py
new file mode 100644
index 00000000..20378cfa
--- /dev/null
+++ b/utils/crawl-locust.py
@@ -0,0 +1,75 @@
+import random
+from locust import HttpUser, between, task
+from bs4 import BeautifulSoup
+
+# To run: python3 -m locust -f utils/crawl-locust.py
+# Dynamically querying projects and versions is not implemented, remember to replace
+# data in `projects` key with a subset of data from the tested instance.
+
+class ElixirUser(HttpUser):
+    wait_time = between(1, 10)
+    projects = (
+        ("linux", (
+            "v6.9.4",
+            "v6.8",
+            "v6.2",
+            "v5.14.15",
+            "v5.9",
+            "v5.4",
+            "v4.17",
+            "v4.10.11",
+            "v4.6",
+            "v3.15",
+            "v3.5.6",
+            "v3.1"
+        )),
+        ("musl",(
+            "v1.2.5" ,
+        )),
+        ("zephyr",(
+            "v3.7.0",
+            "v3.4.0",
+            "v3.0.0",
+            "v2.7.0",
+            "v2.5.0",
+            "v2.3.0",
+            "v1.12.0",
+            "v1.5.0",
+        )),
+    )
+
+    def on_start(self):
+        self.wait()
+        self.index_page()
+
+    def parse_tree(self, phtml):
+        links = phtml.find_all('a', class_='tree-icon')
+
+        for link in links:
+            link_url = f"{self.host}{link['href']}"
+            self.urls.append(link_url)
+
+    def parse_html(self, r):
+        phtml = BeautifulSoup(r.content, 'html.parser')
+        tree = phtml.find(class_='lxrtree')
+        if tree is not None:
+            self.parse_tree(phtml)
+        else:
+            idents = phtml.find_all(class_='ident')
+            for i in idents:
+                self.urls.append(f"{self.host}{i['href']}")
+
+    @task(1)
+    def index_page(self):
+        project, versions = random.choice(self.projects)
+        version = random.choice(versions)
+        r = self.client.get(f"{project}/{version}/source")
+        self.urls = []
+        self.parse_html(r)
+
+    @task(100)
+    def load_random_source_page(self):
+        url = random.choice(self.urls)
+        r = self.client.get(url)
+        self.parse_html(r)
+
diff --git a/utils/crawl.py b/utils/crawl.py
new file mode 100644
index 00000000..a93f3511
--- /dev/null
+++ b/utils/crawl.py
@@ -0,0 +1,52 @@
+import time
+import logging
+import sys
+import random
+from bs4 import BeautifulSoup
+import requests
+
+if __name__ == "__main__":
+    if len(sys.argv) != 4:
+        print(f"usage: {sys.argv[0]} url project version")
+        exit(1)
+
+    base_url = sys.argv[1]
+    project = sys.argv[2]
+    version = sys.argv[3]
+
+    first_url = f"{base_url}/{project}/{version}/source"
+    urls_set = set([first_url])
+    urls_list = [first_url]
+
+    while True:
+        url_index = random.randint(0, len(urls_list)-1)
+        url = urls_list.pop(url_index)
+
+        try:
+            req = requests.get(url, timeout=30)
+        except Exception as e:
+            logging.exception("request failed!")
+            time.sleep(1)
+            continue
+
+        if req.status_code != 200:
+            print("===== ERROR", url, req.status_code)
+
+        phtml = BeautifulSoup(req.text, 'html.parser')
+
+        tree = phtml.find(class_='lxrtree')
+        if tree is not None:
+            links = phtml.find_all('a', class_='tree-icon')
+
+            for link in links:
+                link_url = f"{base_url}{link['href']}"
+                if link_url not in urls_set:
+                    urls_set.add(link_url)
+                    urls_list.append(link_url)
+
+            urls_set.remove(url)
+
+        duration = req.elapsed.total_seconds()
+        print(url, req.status_code, duration, '' if duration < 1 else 'LONG')
+
+