我们正面临redis的问题，其中redis的'dir'路径在没有任何通知的情况下设置。
导致以下错误（写入redis时）。
MISCONF Redis配置为保存RDB快照，但目前是 无法在磁盘上保留。可以修改数据集的命令是 禁用。请检查Redis日志以获取有关错误的详细信息。
信息我们必须支持这一点。
- Redis配置文件
查看我们的redis.conf（位于/etc/redis/redis.conf），提示我们dir路径值为...
# The working directory.
#
# The DB will be written inside this directory, with the filename specified
# above using the 'dbfilename' configuration directive.
#
# The Append Only File will also be created inside this directory.
#
# Note that you must specify a directory here, not a file name.
dir /var/lib/redis
- Redis INFO结果。
{＆＃34; redis_version＆＃34; =＆gt;＆＃34; 2.8.17＆＃34;，＆＃34; redis_git_sha1＆＃34; =＆gt;＆＃34; 00000000＆＃34;， ＆＃34; redis_git_dirty＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; redis_build_id＆＃34; =＆gt;＆＃34; e2de0295da7f0ee1＆＃34;， ＆＃34; redis_mode＆＃34; =＆gt;＆＃34;独立＆＃34;，＆＃34; os＆＃34; =＆gt;＆＃34; Linux 3.16.0-4-amd64 x86_64＆＃34;， ＆＃34; arch_bits＆＃34; =＆gt;＆＃34; 64＆＃34;，＆＃34; multiplexing_api＆＃34; =＆gt;＆＃34; epoll＆＃34;， ＆＃34; gcc_version＆＃34; =＆gt;＆＃34; 4.9.2＆＃34;，＆＃34; process_id＆＃34; =＆gt;＆＃34; 6860＆＃34;， ＆＃34; run_id＆＃34; = GT;＆＃34; 4ecd0ff19798a88793e7a3ce281f398ba9aa2f5a＆＃34 ;, ＆＃34; tcp_port＆＃34; =＆gt;＆＃34; 6379＆＃34;，＆＃34; uptime_in_seconds＆＃34; =＆gt;＆＃34; 99525＆＃34;， ＆＃34; uptime_in_days＆＃34; =＆gt;＆＃34; 1＆＃34;，＆＃34; hz＆＃34; =＆gt;＆＃34; 10＆＃34;，＆＃34; lru_clock＆＃34; =＆gt; ;＆＃34; 11741717＆＃34 ;, ＆＃34; config_file＆＃34; =＆gt;＆＃34; /etc/redis/redis.conf" ;,＆＃34; connected_clients＆＃34; =＆gt;＆＃34; 2＆＃34;， ＆＃34; client_longest_output_list＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; client_biggest_input_buf＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; blocked_clients＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; used_memory＆＃34; =＆gt;＆＃34; 784840＆＃34;， ＆＃34; used_memory_human＆＃34; =＆gt;＆＃34; 766.45K＆＃34;，＆＃34; used_memory_rss＆＃34; =＆gt;＆＃34; 3624960＆＃34;， ＆＃34; used_memory_peak＆＃34; =＆gt;＆＃34; 784840＆＃34;，＆＃34; used_memory_peak_human＆＃34; =＆gt;＆＃34; 766.45K＆＃34;， ＆＃34; used_memory_lua＆＃34; =＆gt;＆＃34; 33792＆＃34;，＆＃34; mem_fragmentation_ratio＆＃34; =＆gt;＆＃34; 4.62＆＃34;， ＆＃34; mem_allocator＆＃34; =＆gt;＆＃34; jemalloc-3.6.0＆＃34;，＆＃34; loading＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rdb_changes_since_last_save＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; rdb_bgsave_in_progress＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rdb_last_save_time＆＃34; =＆gt;＆＃34; 1454582288＆＃34;，＆＃34; rdb_last_bgsave_status＆＃34; =＆gt;＆＃34; ok＆＃34;， ＆＃34; rdb_last_bgsave_time_sec＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; rdb_current_bgsave_time_sec＆＃34; =＆gt;＆＃34; -1＆＃34;， ＆＃34; aof_enabled＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; aof_rewrite_in_progress＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; aof_rewrite_scheduled＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; aof_last_rewrite_time_sec＆＃34; =＆gt;＆＃34; -1＆＃34;， ＆＃34; aof_current_rewrite_time_sec＆＃34; = GT;＆＃34; -1＆＃34 ;, ＆＃34; aof_last_bgrewrite_status＆＃34; =＆gt;＆＃34; ok＆＃34;，＆＃34; aof_last_write_status＆＃34; =＆gt;＆＃34; ok＆＃34;， ＆＃34; total_connections_received＆＃34; = GT;＆＃34; 12＆＃34 ;, ＆＃34; total_commands_processed＆＃34; =＆gt;＆＃34; 4651＆＃34;，＆＃34; instantaneous_ops_per_sec＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rejected_connections＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; sync_full＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; sync_partial_ok＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; sync_partial_err＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; expired_keys＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; evicted_keys＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; keyspace_hits＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; keyspace_misses＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; pubsub_channels＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; pubsub_patterns＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; latest_fork_usec＆＃34; =＆gt;＆＃34; 206＆＃34;，＆＃34;角色＆＃34; =＆gt; ;＆＃34;主＆＃34 ;, ＆＃34; connected_slaves＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; master_repl_offset＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; repl_backlog_active＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; repl_backlog_size＆＃34; =＆gt;＆＃34; 1048576＆＃34;， ＆＃34; repl_backlog_first_byte_offset＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; repl_backlog_histlen＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; used_cpu_sys＆＃34; =＆gt;＆＃34; 57.78＆＃34;，＆＃34; used_cpu_user＆＃34; =＆gt;＆＃34; 25.71＆＃34;， ＆＃34; used_cpu_sys_children＆＃34; =＆gt;＆＃34; 0.00＆＃34;，＆＃34; used_cpu_user_children＆＃34; =＆gt;＆＃34; 0.00＆＃34;， ＆＃34; DB0＆＃34; = GT;＆＃34;键= 2851，到期= 0，avg_ttl = 0＆＃34;}
关键点
redis.info['config_file']
=> /etc/redis/redis.conf
Redis的进程ID
redis.info['process_id']
=> 6860
- Redis配置
然后我检查了redis config和......
redis.config "get" ,"dir"
=> ["dir", "/etc/ssh"]
## ideally should be /var/lib/redis
事情，我做了修复上述错误。
redis.config "set" ,"dir","/var/lib/redis"
=> "OK"
redis.config "get", "dir"
=> ["dir", "/var/lib/redis"]
这种解决方法（上述问题）。但对于临时目的只是因为......
几个小时后我注意到了这一点。
redis.config "get" ,"dir"
=> ["dir", "/etc/ssh"]
再次检查redis配置
- Redis信息
{＆＃34; redis_version＆＃34; =＆gt;＆＃34; 2.8.17＆＃34;，＆＃34; redis_git_sha1＆＃34; =＆gt;＆＃34; 00000000＆＃34;， ＆＃34; redis_git_dirty＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; redis_build_id＆＃34; =＆gt;＆＃34; e2de0295da7f0ee1＆＃34;， ＆＃34; redis_mode＆＃34; =＆gt;＆＃34;独立＆＃34;，＆＃34; os＆＃34; =＆gt;＆＃34; Linux 3.16.0-4-amd64 x86_64＆＃34;， ＆＃34; arch_bits＆＃34; =＆gt;＆＃34; 64＆＃34;，＆＃34; multiplexing_api＆＃34; =＆gt;＆＃34; epoll＆＃34;， ＆＃34; gcc_version＆＃34; =＆gt;＆＃34; 4.9.2＆＃34;，＆＃34; process_id＆＃34; =＆gt;＆＃34; 6860＆＃34;， ＆＃34; run_id＆＃34; = GT;＆＃34; 4ecd0ff19798a88793e7a3ce281f398ba9aa2f5a＆＃34 ;, ＆＃34; tcp_port＆＃34; =＆gt;＆＃34; 6379＆＃34;，＆＃34; uptime_in_seconds＆＃34; =＆gt;＆＃34; 109032＆＃34;， ＆＃34; uptime_in_days＆＃34; =＆gt;＆＃34; 1＆＃34;，＆＃34; hz＆＃34; =＆gt;＆＃34; 10＆＃34;，＆＃34; lru_clock＆＃34; =＆gt; ;＆＃34; 11751224＆＃34 ;, ＆＃34; config_file＆＃34; =＆gt;＆＃34; /etc/redis/redis.conf" ;,＆＃34; connected_clients＆＃34; =＆gt;＆＃34; 2＆＃34;， ＆＃34; client_longest_output_list＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; client_biggest_input_buf＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; blocked_clients＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; used_memory＆＃34; =＆gt;＆＃34; 788984＆＃34;， ＆＃34; used_memory_human＆＃34; =＆gt;＆＃34; 770.49K＆＃34;，＆＃34; used_memory_rss＆＃34; =＆gt;＆＃34; 3629056＆＃34;， ＆＃34; used_memory_peak＆＃34; =＆gt;＆＃34; 808952＆＃34;，＆＃34; used_memory_peak_human＆＃34; =＆gt;＆＃34; 789.99K＆＃34;， ＆＃34; used_memory_lua＆＃34; =＆gt;＆＃34; 33792＆＃34;，＆＃34; mem_fragmentation_ratio＆＃34; =＆gt;＆＃34; 4.60＆＃34;， ＆＃34; mem_allocator＆＃34; =＆gt;＆＃34; jemalloc-3.6.0＆＃34;，＆＃34; loading＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rdb_changes_since_last_save＆＃34; =＆gt;＆＃34; 82＆＃34;，＆＃34; rdb_bgsave_in_progress＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rdb_last_save_time＆＃34; =＆gt;＆＃34; 1454591521＆＃34;，＆＃34; rdb_last_bgsave_status＆＃34; =＆gt;＆＃34; ok＆＃34;， ＆＃34; rdb_last_bgsave_time_sec＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; rdb_current_bgsave_time_sec＆＃34; =＆gt;＆＃34; -1＆＃34;， ＆＃34; aof_enabled＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; aof_rewrite_in_progress＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; aof_rewrite_scheduled＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; aof_last_rewrite_time_sec＆＃34; =＆gt;＆＃34; -1＆＃34;， ＆＃34; aof_current_rewrite_time_sec＆＃34; = GT;＆＃34; -1＆＃34 ;, ＆＃34; aof_last_bgrewrite_status＆＃34; =＆gt;＆＃34; ok＆＃34;，＆＃34; aof_last_write_status＆＃34; =＆gt;＆＃34; ok＆＃34;， ＆＃34; total_connections_received＆＃34; = GT;＆＃34; 14＆＃34 ;, ＆＃34; total_commands_processed＆＃34; =＆gt;＆＃34; 6966＆＃34;，＆＃34; instantaneous_ops_per_sec＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; rejected_connections＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; sync_full＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; sync_partial_ok＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; sync_partial_err＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; expired_keys＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; evicted_keys＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; keyspace_hits＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; keyspace_misses＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; pubsub_channels＆＃34; =＆gt; ;＆＃34; 0＆＃34 ;, ＆＃34; pubsub_patterns＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; latest_fork_usec＆＃34; =＆gt;＆＃34; 135＆＃34;，＆＃34;角色＆＃34; =＆gt; ;＆＃34;主＆＃34 ;, ＆＃34; connected_slaves＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; master_repl_offset＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; repl_backlog_active＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; repl_backlog_size＆＃34; =＆gt;＆＃34; 1048576＆＃34;， ＆＃34; repl_backlog_first_byte_offset＆＃34; =＆gt;＆＃34; 0＆＃34;，＆＃34; repl_backlog_histlen＆＃34; =＆gt;＆＃34; 0＆＃34;， ＆＃34; used_cpu_sys＆＃34; =＆gt;＆＃34; 62.96＆＃34;，＆＃34; used_cpu_user＆＃34; =＆gt;＆＃34; 27.97＆＃34;， ＆＃34; used_cpu_sys_children＆＃34; =＆gt;＆＃34; 0.00＆＃34;，＆＃34; used_cpu_user_children＆＃34; =＆gt;＆＃34; 0.00＆＃34;， ＆＃34; DB0＆＃34; = GT;＆＃34;键= 2902，到期= 0，avg_ttl = 0＆＃34;}
关键点。
redis.info["process_id"]
=> 6860
redis.info['config_file']
=> /etc/redis/redis.conf
所以，
Redis Server从未重启过。 （process_id对此有很多说法。）
Redis正在选择正确的配置文件，即/etc/redis/redis.conf。
那么，如果第1点和第2点都是真的，那么导致redis在几小时后将config dir值设置为/etc/ssh的原因是什么？ （我从昨天开始重置配置5-6次。）
注意：
还重新启动了服务器（昨天），希望永久解决问题。但没有运气。
注意：
回应@Keith回答
是的！服务器完全支持限制性防火墙。
思想？