gortiz commented on code in PR #13303:
URL: https://github.com/apache/pinot/pull/13303#discussion_r1722904792


##########
pinot-segment-spi/src/main/java/org/apache/pinot/segment/spi/memory/PagedPinotOutputStream.java:
##########
@@ -291,35 +290,8 @@ public void close()
   }
 
   public static abstract class PageAllocator {
-    public static final int MIN_RECOMMENDED_PAGE_SIZE;
-    public static final int MAX_RECOMMENDED_PAGE_SIZE;
-
-    static {
-      int minRecommendedPageSize = -1;
-      int maxRecommendedPageSize = -1;
-      try {
-        switch (ArchUtils.getProcessor().getType()) {
-          case AARCH_64:
-            // ARM processors support 4KB and 1MB pages
-            minRecommendedPageSize = 16 * 1024;
-            maxRecommendedPageSize = 1024 * 1024;
-            break;
-          case X86:
-          default:
-            // X86 processors support 4KB and 4MB pages
-            minRecommendedPageSize = 4 * 1024;
-            maxRecommendedPageSize = 4 * 1024 * 1024;
-            break;
-        }
-      } catch (Throwable t) {
-        LOGGER.warn("Could not determine processor architecture. Falling back 
to default values", t);
-        // Fallback to 4KB and 4MBs
-        minRecommendedPageSize = 4 * 1024;
-        maxRecommendedPageSize = 4 * 1024 * 1024;
-      }
-      MIN_RECOMMENDED_PAGE_SIZE = minRecommendedPageSize;
-      MAX_RECOMMENDED_PAGE_SIZE = maxRecommendedPageSize;
-    }
+    public static final int MIN_RECOMMENDED_PAGE_SIZE = 16 * 1024;
+    public static final int MAX_RECOMMENDED_PAGE_SIZE = 1024 * 1024;

Review Comment:
   No, we are not using DirectPageAllocator. I've asked Cliff Click about 
whether direct byte buffers should be slower than heap ones and he didn't find 
any reason why that would happen, so probably there is something wrong in the 
way we are using them. In the future we should dedicate some time to find the 
reason why direct bytebuffers are slower in this benchmark



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org
For additional commands, e-mail: commits-h...@pinot.apache.org

Reply via email to