请教一下，在部署fastdfs的storage碰到了问题

Question

log日志文件的提示如下：

INFO - FastDFS v1.19, base_path=/data/fastdfs/data, store_path_count=1, subdir_count_per_path=256, group_name=group1, network_timeout=60s, port=23000, bind_addr=, max_connections=2048, heart_beat_interval=30s, stat_report_interval=60s, tracker_server_count=1, sync_wait_msec=200ms, sync_interval=0ms, sync_start_time=00:00, sync_end_time: 23:59, allow_ip_count=-1, file_distribute_path_mode=0, file_distribute_rotate_count=100, fsync_after_written_bytes=0, sync_log_buff_interval=10s, sync_binlog_buff_interval=60s, check_file_duplicate=0, FDHT group count=0, FDHT server count=0, FDHT key_namespace=, FDHT keep_alive=0
[2009-08-03 20:08:44] INFO - file: tracker_client_thread.c, line: 191, successfully connect to tracker server 192.168.158.91:22122
[2009-08-03 20:08:44] ERROR - file: ../common/shared_func.c, line: 1635, create thread failed, startup threads: 303, errno: 12, error info: Cannot allocate memory
[2009-08-03 20:08:44] ERROR - file: storage_sync.c, line: 1940, create thread failed, errno: 12, error info: Cannot allocate memory

另外一台storage一样的编译，就没出现这个问题，这台是centos 5.3 64位系统

出问题的系统是 centos 5.3 32位系统

[ 本帖最后由 chhxo 于 2009-8-3 21:12 编辑 ]

Answer 1

线程数超过了系统上限。
可以将其改小，比如改为256。
配置文件中的参数名为max_connections

Answer 2

用ulimit -s 1024改小了系统的线程栈大小，这样就能提升系统线程数上限了

Answer 3

谢谢LZ提供的解决方案。
FastDFS V1.21支持配置线程栈大小，缺省值为1MB。