Description of problem: Glusterfs version: 3.3.0qa45 gluster-swift version: swift-rc1-rpm glusterfs volume type: distribute-replicate(2x2) node_timeout in proxy,containter,object conf files is : 60 worker threads in proxy, container, object, account is : 4 Created 250 objects. Now tried to DELETE while issuing curl command in parallel, some of objects are not removed and for them the below mentioned messages are found in /var/log/messages Jun 5 14:45:44 gqac022 proxy-server 10.16.156.3 10.16.156.3 05/Jun/2012/14/45/44 DELETE /v1/AUTH_test/cont1/f.126 HTTP/1.0 404 - curl/7.19.7%20%28x86_64-redhat-linux-gnu%29%20libcurl/7.19.7%20NSS/3.12.9.0%20zlib/1.2.3%20libidn/1.18%20libssh2/1.2.2 test%2CAUTH_tk97edb64f5c034fb08097c94b27da92e2 - - - txb509d2620aff4c93955aa2a342e26916 - 24.4095 - Jun 5 14:45:45 gqac022 container-server 127.0.0.1 - - [05/Jun/2012:14:45:45 +0000] "HEAD /sdb1/181472/AUTH_test/cont1" 204 - "tx84541617bbdd43dc85e9920a790a3943" "-" "-" 0.5272 Jun 5 14:45:45 gqac022 object-server STDOUT: mkstemp----- (txn: tx84541617bbdd43dc85e9920a790a3943) Jun 5 14:45:45 gqac022 object-server STDOUT: /mnt/gluster-object/AUTH_test/tmp (txn: tx84541617bbdd43dc85e9920a790a3943) Version-Release number of selected component (if applicable): 3.3.0qa45 How reproducible: always Steps to Reproduce: 1. as mentioned above 2. 3. Actual results: Expected results: Additional info:
Can we test this with rc2 rpms and also by setting object_only=yes?
Does this problem occur even now?
This should be resolved on the latest set of rpm's gluster-swift-1.7.4.