Merge pull request #986 from deisseroth-lab/jobs-error-stack-medium-blob

Changes ~jobs.error_stack from blob to mediumblob

Author: dimitri-yatsenko

CHANGELOG.md

@@ -15,6 +15,7 @@
 * Bugfix - Fix sql code generation to comply with sql mode `ONLY_FULL_GROUP_BY` (#916) PR #965
 * Bugfix - Fix count for left-joined `QueryExpressions` (#951) PR #966
 * Bugfix - Fix assertion error when performing a union into a join (#930) PR #967
+* Update `~jobs.error_stack` from blob to mediumblob to allow error stacks >64kB in jobs (#984) PR #986
 
 ### 0.13.2 -- May 7, 2021
 * Update `setuptools_certificate` dependency to new name `otumat`
@@ -248,7 +249,7 @@ Documentation and tutorials available at https://docs.datajoint.io and https://t
 * ERD() no longer text the context argument.
 * ERD.draw() now takes an optional context argument.  By default uses the caller's locals.
 
-### 0.3.2.   
+### 0.3.2.
 * Fixed issue #223:  `insert` can insert relations without fetching.
 * ERD() now takes the `context` argument, which specifies in which context to look for classes. The default is taken from the argument (schema or relation).
 * ERD.draw() no longer has the `prefix` argument: class names are shown as found in the context.

datajoint/jobs.py

@@ -32,7 +32,7 @@ def __init__(self, conn, database):
         status  :enum('reserved','error','ignore')  # if tuple is missing, the job is available
         key=null  :blob  # structure containing the key
         error_message=""  :varchar({error_message_length})  # error message returned if failed
-        error_stack=null  :blob  # error stack if failed
+        error_stack=null  :mediumblob  # error stack if failed
         user="" :varchar(255) # database user
         host=""  :varchar(255)  # system hostname
         pid=0  :int unsigned  # system process id

docs-parts/intro/Releases_lang1.rst

@@ -14,6 +14,7 @@
 * Bugfix - Fix count for left-joined ``QueryExpressions`` (#951) PR #966
 * Bugfix - Fix assertion error when performing a union into a join (#930) PR #967
 * Bugfix - Fix regression issue with `DISTINCT` clause and `GROUP_BY` (#914) PR #963
+* Update `~jobs.error_stack` from blob to mediumblob to allow error stacks >64kB in jobs (#984) PR #986
 
 0.13.2 -- May 7, 2021
 ----------------------

tests/test_jobs.py

@@ -122,3 +122,23 @@ def test_long_error_message():
     assert_true(error_message == short_error_message, 'error messages do not agree')
     assert_false(error_message.endswith(TRUNCATION_APPENDIX), 'error message should not be truncated')
     schema.schema.jobs.delete()
+
+
+def test_long_error_stack():
+    # clear out jobs table
+    schema.schema.jobs.delete()
+
+    # create long error stack
+    STACK_SIZE = 89942  # Does not fit into small blob (should be 64k, but found to be higher)
+    long_error_stack = ''.join(random.choice(string.ascii_letters) for _ in range(STACK_SIZE))
+    assert subjects
+    table_name = 'fake_table'
+
+    key = subjects.fetch('KEY')[0]
+
+    # test long error stack
+    schema.schema.jobs.reserve(table_name, key)
+    schema.schema.jobs.error(table_name, key, 'error message', long_error_stack)
+    error_stack = schema.schema.jobs.fetch1('error_stack')
+    assert error_stack == long_error_stack, 'error stacks do not agree'
+    schema.schema.jobs.delete()