diff --git a/tests/e2e/libervia-cli/test_libervia-cli.py b/tests/e2e/libervia-cli/test_libervia-cli.py
index 15087df3..7ad41746 100644
--- a/tests/e2e/libervia-cli/test_libervia-cli.py
+++ b/tests/e2e/libervia-cli/test_libervia-cli.py
@@ -21,6 +21,7 @@ import shutil
 import pytest
 import sh
 from sh import li
+from time import sleep
 from sat.tools.common import uri
 
 
@@ -168,19 +169,35 @@ class TestLiberviaCliPubsub:
             title=au_txt,
             syntax="markdown"
         )
-        # we get the blog to be activate the cache for it
+        # we get the blog to activate the cache for it
         li.blog.get(max_items=1)
         # FTS
-        found = li_json.pubsub.cache.search(type="blog", fts='Slovakia OR "New Caledonia"')
+        found = []
+        for __ in range(5):
+            found = li_json.pubsub.cache.search(
+                type="blog", fts='Slovakia OR "New Caledonia"'
+            )
+            if found:
+                break
+            else:
+                # retrieving blog triggers the caching, but it's done in parallel
+                # thus we may have nothing in cache yet
+                sleep(0.5)
         assert len(found) == 2
         assert all(i["content"] in (sk_txt, nc_txt) for i in found)
         # search by field
-        found = li_json.pubsub.cache.search("-F", "tags", "overlap", "travel", type="blog")
+        found = li_json.pubsub.cache.search(
+            "-F", "tags", "overlap", "travel", type="blog"
+        )
         assert len(found) == 4
-        found = li_json.pubsub.cache.search("-F", "tags", "overlap", "europe", type="blog")
+        found = li_json.pubsub.cache.search(
+            "-F", "tags", "overlap", "europe", type="blog"
+        )
         assert len(found) == 2
         assert all(i["content"] in (sk_txt, fr_txt) for i in found)
-        found = li_json.pubsub.cache.search("-F", "tags", "ioverlap", "SOUTH PACIFIC", type="blog")
+        found = li_json.pubsub.cache.search(
+            "-F", "tags", "ioverlap", "SOUTH PACIFIC", type="blog"
+        )
         assert all(i["content"] in (nc_txt, au_txt) for i in found)