(tika) branch main updated: improve flaky tests (#2956)

[tallison]

This is an automated email from the ASF dual-hosted git repository.

tballison pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/tika.git


The following commit(s) were added to refs/heads/main by this push:
     new 63e22d08ef improve flaky tests (#2956)
63e22d08ef is described below

commit 63e22d08ef249cc73a6d02da7bc199fc3623a607
Author: Tim Allison <talli...@apache.org>
AuthorDate: Wed Jul 15 08:09:01 2026 -0400

    improve flaky tests (#2956)
---
 .../pipes/elasticsearch/tests/ElasticsearchTest.java | 19 +++++++++++--------
 .../tika/pipes/opensearch/tests/OpenSearchTest.java  | 20 ++++++++++++--------
 2 files changed, 23 insertions(+), 16 deletions(-)

diff --git 
a/tika-integration-tests/tika-pipes-es-integration-tests/src/test/java/org/apache/tika/pipes/elasticsearch/tests/ElasticsearchTest.java
 
b/tika-integration-tests/tika-pipes-es-integration-tests/src/test/java/org/apache/tika/pipes/elasticsearch/tests/ElasticsearchTest.java
index 7d43ce71ee..571cbaf9a0 100644
--- 
a/tika-integration-tests/tika-pipes-es-integration-tests/src/test/java/org/apache/tika/pipes/elasticsearch/tests/ElasticsearchTest.java
+++ 
b/tika-integration-tests/tika-pipes-es-integration-tests/src/test/java/org/apache/tika/pipes/elasticsearch/tests/ElasticsearchTest.java
@@ -178,18 +178,21 @@ public class ElasticsearchTest {
         assertEquals(1, (int) statusCounts.get("EMIT_SUCCESS"),
                 "should have had 1 emit success: " + statusCounts);
         assertEquals(2, numberOfCrashes(statusCounts),
-                "should have had 2 OOM or 1 OOM and 1 timeout: " + 
statusCounts);
+                "should have had 2 forked-process crashes 
(OOM/TIMEOUT/UNSPECIFIED_CRASH): " +
+                        statusCounts);
     }
 
     private int numberOfCrashes(Map<String, Integer> statusCounts) {
-        Integer oom = statusCounts.get("OOM");
-        Integer timeout = statusCounts.get("TIMEOUT");
+        // oom.xml (a real heap exhaustion) and fake_oom.xml both crash the 
fork; how a genuine OOM
+        // surfaces -- OOM vs UNSPECIFIED_CRASH vs TIMEOUT -- is 
nondeterministic under load, but all
+        // three are PipesResult PROCESS_CRASH statuses. Count the whole 
category so the assertion is
+        // deterministic and doesn't flake on the exact sub-classification.
         int sum = 0;
-        if (oom != null) {
-            sum += oom;
-        }
-        if (timeout != null) {
-            sum += timeout;
+        for (String crashStatus : new String[]{"OOM", "TIMEOUT", 
"UNSPECIFIED_CRASH"}) {
+            Integer cnt = statusCounts.get(crashStatus);
+            if (cnt != null) {
+                sum += cnt;
+            }
         }
         return sum;
     }
diff --git 
a/tika-integration-tests/tika-pipes-opensearch-integration-tests/src/test/java/org/apache/tika/pipes/opensearch/tests/OpenSearchTest.java
 
b/tika-integration-tests/tika-pipes-opensearch-integration-tests/src/test/java/org/apache/tika/pipes/opensearch/tests/OpenSearchTest.java
index 1e33c22383..745654fc76 100644
--- 
a/tika-integration-tests/tika-pipes-opensearch-integration-tests/src/test/java/org/apache/tika/pipes/opensearch/tests/OpenSearchTest.java
+++ 
b/tika-integration-tests/tika-pipes-opensearch-integration-tests/src/test/java/org/apache/tika/pipes/opensearch/tests/OpenSearchTest.java
@@ -155,19 +155,23 @@ public class OpenSearchTest {
         assertEquals(1, (int) 
statusCounts.get("PARSE_SUCCESS_WITH_EXCEPTION"), "should have had 1 parse 
exception: " + statusCounts);
         //the embedded docx is emitted directly
         assertEquals(1, (int) statusCounts.get("EMIT_SUCCESS"), "should have 
had 1 emit success: " + statusCounts);
-        assertEquals(2, numberOfCrashes(statusCounts), "should have had 2 OOM 
or 1 OOM and 1 timeout: " + statusCounts);
+        assertEquals(2, numberOfCrashes(statusCounts),
+                "should have had 2 forked-process crashes 
(OOM/TIMEOUT/UNSPECIFIED_CRASH): " +
+                        statusCounts);
 
     }
 
     private int numberOfCrashes(Map<String, Integer> statusCounts) {
-        Integer oom = statusCounts.get("OOM");
-        Integer timeout = statusCounts.get("TIMEOUT");
+        // oom.xml (a real heap exhaustion) and fake_oom.xml both crash the 
fork; how a genuine OOM
+        // surfaces -- OOM vs UNSPECIFIED_CRASH vs TIMEOUT -- is 
nondeterministic under load, but all
+        // three are PipesResult PROCESS_CRASH statuses. Count the whole 
category so the assertion is
+        // deterministic and doesn't flake on the exact sub-classification.
         int sum = 0;
-        if (oom != null) {
-            sum += oom;
-        }
-        if (timeout != null) {
-            sum += timeout;
+        for (String crashStatus : new String[]{"OOM", "TIMEOUT", 
"UNSPECIFIED_CRASH"}) {
+            Integer cnt = statusCounts.get(crashStatus);
+            if (cnt != null) {
+                sum += cnt;
+            }
         }
         return sum;
     }