[问题描述]:挂载GlusterFS分布式文件系统报错: [root@qibo-anck014-an23-g6r-1 ~]# mount.glusterfs node1:dis-volume /data/ grep: warning: stray \ before - Fatal glibc error: malloc assertion failure in sysmalloc: (old_top == initial_top (av) && old_size == 0) || ((unsigned long) (old_size) >= MINSIZE && prev_inuse (old_top) && ((unsigned long) old_end & (pagesize - 1)) == 0) /usr/sbin/mount.glusterfs: line 125: 795870 Aborted (core dumped) $cmd_line Mounting glusterfs on /data/ failed. [环境信息]: 机器类型:ECS [内核信息]: 客户端(挂载端) [root@qibo-anck014-an23-g6r-1 ~]# uname -r 5.10.134-14_rc2.1.an23.aarch64 服务端1(部署lusterFS): [root@iZbp14fzsphchtmh4wcau9Z ~]# uname -r 5.10.134-13.2_alpha1.an23.x86_64 服务端2(部署lusterFS): [root@iZbp14fzsphchtmh4wcauaZ ~]# uname -r 5.10.134-13.2_alpha1.an23.x86_64 [操作系统信息]: 客户端(挂载端) [root@qibo-anck014-an23-g6r-1 ~]# cat /etc/os-release NAME="Anolis OS" VERSION="23" ID="anolis" VERSION_ID="23" PLATFORM_ID="platform:an23" PRETTY_NAME="Anolis OS 23" ANSI_COLOR="0;31" HOME_URL="https://openanolis.cn/" BUG_REPORT_URL="https://bugzilla.openanolis.cn/" 服务端1(部署lusterFS): [root@iZbp14fzsphchtmh4wcau9Z ~]# uname -r 5.10.134-13.2_alpha1.an23.x86_64 [root@iZbp14fzsphchtmh4wcau9Z ~]# cat /etc/os-release NAME="Anolis OS" VERSION="23" ID="anolis" VERSION_ID="23" PLATFORM_ID="platform:an23" PRETTY_NAME="Anolis OS 23" ANSI_COLOR="0;31" HOME_URL="https://openanolis.cn/" BUG_REPORT_URL="https://bugzilla.openanolis.cn/" 服务端2(部署lusterFS): [root@iZbp14fzsphchtmh4wcauaZ ~]# cat /etc/os-release NAME="Anolis OS" VERSION="23" ID="anolis" VERSION_ID="23" PLATFORM_ID="platform:an23" PRETTY_NAME="Anolis OS 23" ANSI_COLOR="0;31" HOME_URL="https://openanolis.cn/" BUG_REPORT_URL="https://bugzilla.openanolis.cn/" [问题发生概率]:必现 [复现步骤]: 前置条件(安装部署glusterd): 1、客户端执行:yum install glusterfs glusterfs-fuse安装软件包; 2、服务端执行:yum -y install glusterfs glusterfs-server glusterfs-fuse glusterfs-rdma安装软件包 3、所有机器配置/etc/hosts文件类型如下: Server1_ip node1 如:192.168.1.1 node1 Server2_ip node2 4、服务端磁盘挂载,如将/dev/vdb1挂载到/data目录下: mount /dev/vdb1 /data/ 5、服务端客户端执行systemctl restart glusterd重启服务。 测试步骤(创建与挂载): 1、服务端执行下列命令添加节点创建集群(在一台机器执行即可,无需执行本机节点): gluster peer probe node1 2、服务端创建分布式卷(一台机器执行即可): gluster volume create dis-volume node1:/data node2:/data force 3、服务端执行gluster volume list命令,查看卷列表; 4、服务端执行gluster volume start dis-volume启动新建卷; 5、服务端执行gluster volume info dis-volume命令,查看分布式卷信息; 6、客户端创建挂载目录如:mkdir /data; 7、客户端执行mount.glusterfs node1:dis-volume /data命令,挂载分布式文件系统。
Anolis8会挂载失败,查询了一下原因可能是gluster版本过低: 执行命令: [2023-03-22 16:37:34] [root@iZbp11lagjmwtlkrtvp6hgZ ~]# mount.fluuuse glusterfs node1:dis0-volmenume [2023-03-22 16:38:04] ERROR: Server name/volume name unspecified cannot proceed further.. [2023-03-22 16:38:04] Please specify correct format [2023-03-22 16:38:04] Usage: [2023-03-22 16:38:04] man 8 /usr/sbin/mount.glusterfs [2023-03-22 16:38:04] [root@iZbp11lagjmwtlkrtvp6hgZ ~]# mkdir /opt/test [2023-03-22 16:38:15] [root@iZbp11lagjmwtlkrtvp6hgZ ~]# mkdir /opt/testount.glusterfs node1:dis-volume /opt/test/ [2023-03-22 16:38:22] Mounting glusterfs on /opt/test/ failed. [2023-03-22 16:39:04] [root@iZbp11lagjmwtlkrtvp6hgZ ~]# catac /etc/varvar/log/messages [2023-03-22 16:39:26] Mar 22 16:39:04 iZbp11lagjmwtlkrtvp6hgZ systemd[1]: opt-test.mount: Succeeded. [2023-03-22 16:39:26] Mar 22 16:38:22 iZbp11lagjmwtlkrtvp6hgZ systemd-udevd[531]: Network interface NamePolicy= disabled on kernel command line, ignoring. [2023-03-22 16:39:26] Mar 22 16:37:34 iZbp11lagjmwtlkrtvp6hgZ systemd[1]: Started Dynamically Generate Message Of The Day. [2023-03-22 16:39:26] Mar 22 16:37:33 iZbp11lagjmwtlkrtvp6hgZ systemd[1]: Starting Dynamically Generate Message Of The Day... 日志查看: [root@iZbp11lagjmwtlkrtvp6hgZ ~]# cat /var/log/glusterfs/opt-test.log [2023-03-22 08:40:02.827114] I [MSGID: 100030] [glusterfsd.c:2868:main] 0-/usr/sbin/glusterfs: Started running /usr/sbin/glusterfs version 6.0 (args: /usr/sbin/glusterfs --process-name fuse --volfile-server=node1 --volfile-id=dis-volume /opt/test) [2023-03-22 08:40:02.827646] I [glusterfsd.c:2577:daemonize] 0-glusterfs: Pid of current running process is 32735 [2023-03-22 08:40:02.832939] I [MSGID: 101190] [event-epoll.c:688:event_dispatch_epoll_worker] 0-epoll: Started thread with index 1 [2023-03-22 08:40:02.833171] I [MSGID: 101190] [event-epoll.c:688:event_dispatch_epoll_worker] 0-epoll: Started thread with index 0 [2023-03-22 08:40:45.085347] I [glusterfsd-mgmt.c:2451:mgmt_rpc_notify] 0-glusterfsd-mgmt: disconnected from remote-host: node1 [2023-03-22 08:40:45.085378] I [glusterfsd-mgmt.c:2471:mgmt_rpc_notify] 0-glusterfsd-mgmt: Exhausted all volfile servers [2023-03-22 08:40:45.085531] W [glusterfsd.c:1583:cleanup_and_exit] (-->/lib64/libgfrpc.so.0(+0xf953) [0x7f6e66c0f953] -->/usr/sbin/glusterfs(+0x137c7) [0x5556742917c7] -->/usr/sbin/glusterfs(cleanup_and_exit+0x58) [0x5556742865c8] ) 0-: received signum (1), shutting down [2023-03-22 08:40:45.085553] I [fuse-bridge.c:6966:fini] 0-fuse: Unmounting '/opt/test'. [2023-03-22 08:40:45.085753] I [fuse-bridge.c:6971:fini] 0-fuse: Closing fuse connection to '/opt/test'.
11.0 是 glusterfs 目前的最新版本。 看起来报错和这个很像: https://bugzilla.suse.com/show_bug.cgi?id=1209702
从编译结果上来说,x86-64的版本,有到tcmalloc_minimal的引用,应该不会报错,aarch64 没有链接到tcmalloc_minimal,可能会报错