diff --git a/apps/resm/tasks.py b/apps/resm/tasks.py
index fb2f39b..e0b5a52 100644
--- a/apps/resm/tasks.py
+++ b/apps/resm/tasks.py
@@ -11,6 +11,7 @@ import requests
 from lxml import etree
 from django.conf import settings
 import os
+from celery import current_app
 
 config.email = "caoqianming@foxmail.com"
 config.max_retries = 0
@@ -61,11 +62,11 @@ def get_paper_meta_from_openalex(publication_year:int, search_key:str):
 
 ELSEVIER_APIKEY = 'aa8868cac9e27d6153ab0a0acd7b50bf'
 @shared_task(base=CustomTask)
-def get_abstract_from_elsevier(publication_year: int, number_of_task:int = 100):
-    qs = Paper.objects.filter(
-        publication_year=publication_year,
-        has_abstract=False
-    ).exclude(
+def get_abstract_from_elsevier(publication_year: int = None, number_of_task:int = 100):
+    qs = Paper.objects.filter(has_abstract=True)
+    if publication_year is not None:
+        qs = qs.filter(publication_year=publication_year)
+    qs = qs.exclude(
         fail_reason="elsevier_doi_not_found"
     ).order_by("publication_date")
 
@@ -77,70 +78,84 @@ def get_abstract_from_elsevier(publication_year: int, number_of_task:int = 100):
         "httpAccept": "text/xml"
     }
     err_msg = ""
-    req = requests.Session()
-    for paper in qs[:number_of_task]:
-        try:
-            res = req.get(
-                f"https://api.elsevier.com/content/article/doi/{paper.doi}",
-                params=params,
-                timeout=(3, 15)
-            )
-        except requests.RequestException:
-            err_msg = "elsevier_request_error"
-            break
-
-        if res.status_code == 200:
-            xml_str = res.text
+    if number_of_task is None:
+        papers = qs.all()
+    else:
+        papers = qs[:number_of_task]
+    with requests.Session() as req:
+        for paper in papers:
             try:
-                root = etree.fromstring(xml_str.encode("utf-8"))
-            except etree.XMLSyntaxError:
-                paper.fail_reason = "elsevier_xml_error"
-                paper.save(update_fields=["fail_reason"])
-                continue
-
-            ns = {"dc": "http://purl.org/dc/elements/1.1/",
-            "ce": "http://www.elsevier.com/xml/common/dtd"}
-            abstract = root.xpath("//dc:description/text()", namespaces=ns)
-            if abstract:
-                PaperAbstract.objects.update_or_create(
-                    paper=paper,
-                    defaults={
-                        "abstract": abstract[0].strip(),
-                        "source": "elsevier"
-                    }
+                res = req.get(
+                    f"https://api.elsevier.com/content/article/doi/{paper.doi}",
+                    params=params,
+                    timeout=(3, 15)
                 )
-                paper.has_abstract = True
-                paper.has_abstract_xml = True
+            except requests.RequestException:
+                err_msg = "elsevier_request_error"
+                break
 
-            paras = root.xpath("//ce:para", namespaces=ns)
-            has_fulltext = len(paras) > 0
-            if has_fulltext:
-                paper.has_fulltext = True
-                paper.has_fulltext_xml = True
-                
-            publication_date = paper.publication_date
-            paper_dir = os.path.join(
-                settings.BASE_DIR,
-                "media/papers",
-                str(publication_date.year),
-                str(publication_date.month),
-                str(publication_date.day)
-            )
-            os.makedirs(paper_dir, exist_ok=True)
+            if res.status_code == 200:
+                xml_str = res.text
+                try:
+                    root = etree.fromstring(xml_str.encode("utf-8"))
+                except etree.XMLSyntaxError:
+                    paper.fail_reason = "elsevier_xml_error"
+                    paper.save(update_fields=["fail_reason"])
+                    continue
 
-            safe_doi = paper.doi.replace("/", "_")
-            paper_file = os.path.join(paper_dir, f"{safe_doi}.xml")
-            with open(paper_file, "wb") as f:
-                f.write(xml_str.encode("utf-8"))
+                ns = {"dc": "http://purl.org/dc/elements/1.1/",
+                "ce": "http://www.elsevier.com/xml/common/dtd"}
+                abstract = root.xpath("//dc:description/text()", namespaces=ns)
+                if abstract:
+                    PaperAbstract.objects.update_or_create(
+                        paper=paper,
+                        defaults={
+                            "abstract": abstract[0].strip(),
+                            "source": "elsevier"
+                        }
+                    )
+                    paper.has_abstract = True
+                    paper.has_abstract_xml = True
+                    paper.fetch_status = "abstract_ready"
 
-            paper.save(update_fields=["has_abstract", "has_abstract_xml", "has_fulltext", "has_fulltext_xml", "update_time"])
+                paras = root.xpath("//ce:para", namespaces=ns)
+                has_fulltext = len(paras) > 0
+                if has_fulltext:
+                    paper.has_fulltext = True
+                    paper.has_fulltext_xml = True
+                    paper.fetch_status = "fulltext_ready"
+                    
+                publication_date = paper.publication_date
+                paper_dir = os.path.join(
+                    settings.BASE_DIR,
+                    "media/papers",
+                    str(publication_date.year),
+                    str(publication_date.month),
+                    str(publication_date.day)
+                )
+                os.makedirs(paper_dir, exist_ok=True)
 
-        elif res.status_code == 404:
-            paper.fail_reason = "elsevier_doi_not_found"
-            paper.save(update_fields=["fail_reason"])
+                safe_doi = paper.doi.replace("/", "_")
+                paper_file = os.path.join(paper_dir, f"{safe_doi}.xml")
+                with open(paper_file, "wb") as f:
+                    f.write(xml_str.encode("utf-8"))
+
+                paper.save(update_fields=["has_abstract", "has_abstract_xml", "has_fulltext", "has_fulltext_xml", "update_time"])
+
+            elif res.status_code == 404:
+                paper.fail_reason = "elsevier_doi_not_found"
+                paper.save(update_fields=["fail_reason"])
     
     remaining_count = qs.count()
     if remaining_count == 0:
         return "done"
     else:
+        current_app.send_task(
+            "apps.resm.tasks.get_abstract_from_elsevier",
+            kwargs={
+                "publication_year": publication_year,
+                "number_of_task": number_of_task,
+            },
+            countdown=5,
+        )
         return f'{err_msg}, remaining {remaining_count} papers'
\ No newline at end of file