From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.1) Gecko/20020827 Description of problem: A 2cpu smp system exhibits poor scheduling performance when a process is writing to disk via the ext3 file system. To drive the bug, create a process that sequentially writes 3GB of data to a file on disk. Observe that 'cat /proc/meminfo' takes minutes to return while the writes are taking place. Observe that 'ssh' stalls for a long time when logging into the machine from a remote machine. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Create a process that sequentially writes to disk. 2. Run the process. 3. Try doing other work on the system. Actual Results: 'cat /proc/meminfo' took minutes to return. 'ssh' took tens of seconds to log into the system. (iostat reports a write throughput of about 10 to 11 MB/sec.) Expected Results: Locking and scheduling algorithms should be tuned so that long running processes that perform heavy I/O do not make the system unusable by other processes which perform light I/O. Additional info: ext3 file system, dual cpu, 2GB of ram, no swap space.
Created attachment 80706 [details] program to generate an I/O load
Thanks for the bug report. However, Red Hat no longer maintains this version of the product. Please upgrade to the latest version and open a new bug if the problem persists. The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, and if you believe this bug is interesting to them, please report the problem in the bug tracker at: http://bugzilla.fedora.us/