部署3台ceph 节点，如下：[root@ceph-1 ceph]# cat ceph.conf[global]fsid = 835f6d27-8e03-4f7f-8d21-b0a26993b529mon_initial_members = ceph-1mon_host = 172.24.193.62auth_cluster_required = cephxauth_service_required = cephxauth_client_required = cephxmon_allow_pool_delete=trueosd_pool_default_size = 2[root@ceph-1 ~]# ceph -s  cluster:    id:     835f6d27-8e03-4f7f-8d21-b0a26993b529    health: HEALTH_OK  services:    mon: 1 daemons, quorum ceph-1    mgr: ceph-1(active)    osd: 3 osds: 3 up, 3 in  data:    pools:   1 pools, 128 pgs    objects: 510  objects, 1.9 GiB    usage:   6.8 GiB used, 233 GiB / 240 GiB avail    pgs:     128 active+clean[root@ceph-1 ~]# ceph osd status+----+--------+-------+-------+--------+---------+--------+---------+-----------+| id |  host  |  used | avail | wr ops | wr data | rd ops | rd data |   state   |+----+--------+-------+-------+--------+---------+--------+---------+-----------+| 0  | ceph-1 | 2170M | 77.8G |    0   |     0   |    0   |     0   | exists,up || 1  | ceph-2 | 2376M | 77.6G |    0   |     0   |    0   |     0   | exists,up || 2  | ceph-3 | 2456M | 77.5G |    0   |     0   |    0   |     0   | exists,up |+----+--------+-------+-------+--------+---------+--------+---------+-----------+[root@ceph-1 ceph]# ceph auth lsinstalled auth entries:osd.0    key: AQDH9qNjHRvdDBAAIS8klDI+PhUUSAwlxmEhKA==    caps: [mgr] allow profile osd    caps: [mon] allow profile osd    caps: [osd] allow *osd.1    key: AQDf9qNjNSttFRAA28htGbtUWAYBxK9kacVHXQ==    caps: [mgr] allow profile osd    caps: [mon] allow profile osd    caps: [osd] allow *osd.2    key: AQDu9qNj43mtGRAAJ4tAU5Ie9carFkhRcwnMnQ==    caps: [mgr] allow profile osd    caps: [mon] allow profile osd    caps: [osd] allow *client.admin    key: AQC+86NjAev/HRAAD/wXEPaDVlEOWMn0H2x8ng==    caps: [mds] allow *    caps: [mgr] allow *    caps: [mon] allow *    caps: [osd] allow *client.bootstrap-mds    key: AQC+86NjHAIAHhAAm0L2LXEzwRx7ijTBMlsmSQ==    caps: [mon] allow profile bootstrap-mdsclient.bootstrap-mgr    key: AQC+86Nj4xEAHhAADcU3pbq04mk4wK8DxNH6hA==    caps: [mon] allow profile bootstrap-mgrclient.bootstrap-osd    key: AQC+86NjryAAHhAAzCw13oaWl8EvdrVRCqc+wQ==    caps: [mon] allow profile bootstrap-osdclient.bootstrap-rbd    key: AQC+86NjEDAAHhAAWev9IkV4zd6gSoUjEEK2ZA==    caps: [mon] allow profile bootstrap-rbdclient.bootstrap-rgw    key: AQC+86NjXz4AHhAADlfczh0/8voJZuDWkHaJ9w==    caps: [mon] allow profile bootstrap-rgwclient.libvirt    key: AQDULP9jtWHCMxAAfIBCbNynpHADCYMwQr8xpg==    caps: [mon] allow r    caps: [osd] allow class-read object_prefix rbd_children, allow rwx pool=libvirt-poolmgr.ceph-1    key: AQAF9KNj6Io3OBAAX6koyzKcwSn+IL5IemY3xg==    caps: [mds] allow *    caps: [mon] allow profile mgr    caps: [osd] allow *在ceph client 上创建vm ，使用ceph rbd 访问 ceph pool libvirt-pool里的qcow2 镜像：     43     <emulator>/usr/local/bin/qemu-system-x86_64</emulator>     44     <disk type='network' device='disk'>     45       <driver name='qemu' type='raw' cache='writeback'/>     46       <auth username='libvirt'>     47         <secret type='ceph' uuid='8187dd57-7e0d-4a9d-8e5a-fe5234d40e51'/>     48       </auth>     49       <source protocol='rbd' name='libvirt-pool/root-vsys_c1.qcow2'>     50         <host name='172.24.193.62' port='6789'/>     51         <host name='172.24.193.63' port='6789'/>     52         <host name='172.24.193.64' port='6789'/>     53       </source>     54       <target dev='vda' bus='virtio'/>     55       <address type='pci' domain='0x0000' bus='0x00' slot='0x09' function='0x0'/>     56     </disk>在ceph-client 上可以利用 client.libvirt 用户访问 ceph 上的libvirt-pool ，如下：如上图卡了，看样子是找不到启动盘了。。。libvirtd 把 root-vsys_c1 xml 中的 启动 虚拟机的命令如下：[root@ceph-client-65 ceph]# rbd --user libvirt ls libvirt-poolroot-vsys_c1.qcow2而我不通过virsh start xxx 方式启动vm，采取直接qemu 启动，则可以成功进入vm 内部，并正常引导kernel这种方式和virsh 方式的不同点，就是 qemu 没有同ceph 进行cephx 验证吧。本人怀疑是 ceph client 接入ceph 认证导致的，可是我都已经参考了此ceph 官网链接：https://docs.ceph.com/en/quincy/rbd/libvirt/ 反复检查了。在创建vm 的客户端节点上都可以访问ceph 集群内。希望有从事ceph 分布式存储这方面实战经验的高手多多指教下这个现象出现的原因及解决方法。谢谢！

创建minio，没配置访问链接的地方，如何配置外部存储的minio?

存储位置不同，打包到生产后，开发环境配置的资源获取不到，有没有能通过中间层，或者配置参数的方式，打包到生产能正常调用不同存储的资源？

ERROR - 2022-12-07 18:05:29 --> Severity: Warning --> fread(): SSL: Connection reset by peer /dev/third_party/huaweicloud-obs/vendor/guzzlehttp/psr7/src/Stream.php 231ERROR - 2022-12-07 17:32:56 --> Severity: error --> Exception: Argument 1 passed to Obs\ObsClient::Obs\Internal\{closure}() must be an instance of GuzzleHttp\Exception\RequestException, instance of GuzzleHttp\Exception\ConnectException given, called in /dev/third_party/huaweicloud-obs/vendor/guzzlehttp/promises/src/Promise.php on line 204 /dev/third_party/huaweicloud-obs/Obs/Internal/SendRequestTrait.php 640

迁移环境:操作系统:    CentOS Linux release 7.6.1810 (AltArch) Linux ### 4.14.0-115.el7a.0.1.aarch64 #1 SMP Sun Nov 25 20:54:21 UTC 2018 aarch64 aarch64 aarch64 GNU/Linux组件: tfs(开源地址:https://github.com/alibaba/tfs/tree/release-2.2.13)依赖:jemalloc,mysql-connector-c,zlib-devel,libuuid,libtbsys.so,libtbnet.so(后两个的迁移案例:https://bbs.huaweicloud.com/forum/thread-102614-1-1.html)迁移过程:里面的脚本等小错误还是比较多的，但这些都不是重点，稍微修改一下就好(后面附带修改后的完整代码)，来看看一下报错    unknown mnemonic 'lock'  -- 'lock'    unknown mnemonic 'xaddl' -- xaddl x20,[x0]    unknown mnemonic 'lock'  --  'lock'    unknown mnemonic 'xaddl'  --  xaddl x7,[x1]错误类似上一篇https://bbs.huaweicloud.com/forum/thread-102614-1-1.html我们接着看具体的代码:src/common/atomic.h-----------------------------------------------------------------------------------------------------#if defined(__i386__) || defined(__x86_64__)        // Atomic operations that C can't guarantee us.  Useful for    // resource counting etc..    // SMP lock prefix                                                                                                               #ifdef CONFIG_SMP    #define LOCK_PREFIX "lock ; "    #else    #define LOCK_PREFIX ""    #endif    //return: the incremented value;    /// 原子地做 8位，16位，32位，64位的++i的操作    /// 该操作虽然参数和返回值都是无符号型整数，但是一样可以    /// 对有符号型整数做操作，只需要做适当的参数转换即可    /// @param pv 指向操作数的指针    /// @return 操作数加1以后的数值    #ifdef __x86_64__    static __inline__ uint64_t atomic_inc(volatile uint64_t * pv)    {      register unsigned long __res;      __asm__ __volatile__ (          "movq $1,%0;"          LOCK_PREFIX "xaddq %0,(%1);"          "incq %0"          :"=a" (__res), "=q" (pv): "1" (pv));      return __res;    }    #endif        static __inline__ uint32_t atomic_inc(volatile uint32_t * pv)    {      register unsigned int __res;      __asm__ __volatile__ (          "movl $1,%0;"          LOCK_PREFIX "xaddl %0,(%1);"          "incl %0"          :"=a" (__res), "=q" (pv): "1" (pv));      return __res;    }        static __inline__ uint16_t atomic_inc(volatile uint16_t * pv)    {      register unsigned short __res;      __asm__ __volatile__ (          "movw $1,%0;"          LOCK_PREFIX "xaddw %0,(%1);"          "incw %0"          :"=a" (__res), "=q" (pv): "1" (pv));      return __res;        }        static __inline__ uint8_t  atomic_inc(volatile uint8_t * pv)    {      register unsigned char __res;      __asm__ __volatile__ (          "movb $1,%0;"          LOCK_PREFIX "xaddb %0,(%1);"          "incb %0"          :"=a" (__res), "=q" (pv): "1" (pv));      return __res;    }        //return: the decremented value;    /// 原子地做 8位，16位，32位，64位的--i的操作    /// 该操作虽然参数和返回值都是无符号型整数，但是一样可以    /// 对有符号型整数做操作，只需要做适当的参数转换即可    /// @param pv 指向操作数的指针    /// @return 操作数减1后的数值    #ifdef __x86_64__    static __inline__ uint64_t atomic_dec(volatile uint64_t * pv)    {      register unsigned long __res;      __asm__ __volatile__ (          "movq $0xffffffffffffffff,%0;"          LOCK_PREFIX "xaddq %0,(%1);"          "decq %0"          : "=a" (__res), "=q" (pv): "1" (pv));      return __res;        }    #endif    static __inline__ uint32_t atomic_dec(volatile uint32_t * pv)    {      register unsigned int __res;      __asm__ __volatile__ (          "movl $0xffffffff,%0;"          LOCK_PREFIX "xaddl %0,(%1);"          "decl %0"          : "=a" (__res), "=q" (pv): "1" (pv));      return __res;        }    static __inline__ uint16_t atomic_dec(volatile uint16_t * pv)    {      register unsigned short __res;      __asm__ __volatile__ (          "movw $0xffff,%0;"          LOCK_PREFIX "xaddw %0,(%1);"          "decw %0"          : "=a" (__res), "=q" (pv): "1" (pv));      return __res;    }        static __inline__ uint8_t  atomic_dec(volatile uint8_t * pv)    {      register unsigned char __res;      __asm__ __volatile__ (          "movb $0xff,%0;"          LOCK_PREFIX "xaddb %0,(%1);"          "decb %0"          : "=a" (__res), "=q" (pv): "1" (pv));      return __res;    }        //return: the initial value of *pv    /// 原子地做 8位，16位，32位，64位的加法的操作    /// 该操作虽然参数和返回值都是无符号型整数，但是一样可以    /// 对有符号型整数做操作，只需要做适当的参数转换即可    /// @param pv 指向操作数的指针    /// @return 操作数加法之前的数值    #ifdef __x86_64__    static __inline__ uint64_t atomic_add(volatile uint64_t * pv, const uint64_t av)    {      //:"=a" (__res), "=q" (pv): "m"(av), "1" (pv));      register unsigned long __res;      __asm__ __volatile__ (          "movq %2,%0;"          LOCK_PREFIX "xaddq %0,(%1);"          :"=a" (__res), "=q" (pv): "mr"(av), "1" (pv));      return __res;    }    #endif    static __inline__ uint32_t atomic_add(volatile uint32_t * pv, const uint32_t av)    {      //:"=a" (__res), "=q" (pv): "m"(av), "1" (pv));      register unsigned int __res;      __asm__ __volatile__ (          "movl %2,%0;"          LOCK_PREFIX "xaddl %0,(%1);"          :"=a" (__res), "=q" (pv): "mr"(av), "1" (pv));      return __res;    }        static __inline__ uint16_t atomic_add(volatile uint16_t * pv, const uint16_t av)    {      //:"=a" (__res), "=q" (pv): "m"(av), "1" (pv));      register unsigned short __res;      __asm__ __volatile__ (          "movw %2,%0;"          LOCK_PREFIX "xaddw %0,(%1);"          :"=a" (__res), "=q" (pv): "mr"(av), "1" (pv));      return __res;    }        static __inline__ uint8_t  atomic_add(volatile uint8_t * pv, const uint8_t av)    {      //:"=a" (__res), "=q" (pv): "m"(av), "1" (pv));      register unsigned char __res;      __asm__ __volatile__ (          "movb %2,%0;"          LOCK_PREFIX "xaddb %0,(%1);"          :"=a" (__res), "=q" (pv): "mr"(av), "1" (pv));      return __res;    }        //function: set *pv to nv    //return: the initial value of *pv    /// 原子地把nv赋值给pv指向的整数，支持8位，16位，32位，84位操作    /// @param pv 待赋值的整数（目的操作数）    /// @param nv 向pv赋的整数    /// @return pv指向的赋值前的数值    #ifdef __x86_64__    static __inline__ uint64_t atomic_exchange(volatile uint64_t * pv, const uint64_t nv)    {      register unsigned long __res;      __asm__ __volatile__ (          "1:"          LOCK_PREFIX "cmpxchgq %3,(%1);"                \          "jne 1b":           "=a" (__res), "=q" (pv): "1" (pv), "q" (nv), "0" (*pv));      return __res;    }    #endif    static __inline__ uint32_t atomic_exchange(volatile uint32_t * pv, const uint32_t nv)    {      register unsigned int __res;      __asm__ __volatile__ (          "1:"          LOCK_PREFIX "cmpxchgl %3,(%1);"                \          "jne 1b":           "=a" (__res), "=q" (pv): "1" (pv), "q" (nv), "0" (*pv));      return __res;    }        static __inline__ uint16_t atomic_exchange(volatile uint16_t * pv, const uint16_t nv)    {      register unsigned short __res;      __asm__ __volatile__ (          "1:"          LOCK_PREFIX "cmpxchgw %3,(%1);"                \          "jne 1b":           "=a" (__res), "=q" (pv): "1" (pv), "q" (nv), "0" (*pv));      return __res;    }        static __inline__ uint8_t  atomic_exchange(volatile uint8_t * pv, const uint8_t nv)    {      register unsigned char __res;      __asm__ __volatile__ (          "1:"          LOCK_PREFIX "cmpxchgb %3,(%1);"                \          "jne 1b":           "=a" (__res), "=q" (pv): "1" (pv), "q" (nv), "0" (*pv));       return __res;    }        //function: compare *pv to cv, if equal, set *pv to nv, otherwise do nothing.    //return: the initial value of *pv    /// 比较pv和cv，如果两者相等，则返回pv原有数值并且把nv赋值给pv    /// 否则什么也不作，返回pv原有数值    /// @param pv 待赋值的整数（目的操作数）    /// @param nv 向pv赋的整数    /// @param cv 和pv比较的整数    /// @return pv指向的操作前的数值    #ifdef __x86_64__    static __inline__ uint64_t atomic_compare_exchange(volatile uint64_t * pv,         const uint64_t nv, const uint64_t cv)    {      register unsigned long __res;      __asm__ __volatile__ (          LOCK_PREFIX "cmpxchgq %3,(%1)"          : "=a" (__res), "=q" (pv) : "1" (pv), "q" (nv), "0" (cv));      return __res;    }    #endif    static __inline__ uint32_t atomic_compare_exchange(volatile uint32_t * pv,         const uint32_t nv, const uint32_t cv)    {      register unsigned int __res;      __asm__ __volatile__ (          LOCK_PREFIX "cmpxchgl %3,(%1)"          : "=a" (__res), "=q" (pv) : "1" (pv), "q" (nv), "0" (cv));      return __res;    }    static __inline__ uint16_t atomic_compare_exchange(volatile uint16_t * pv,         const uint16_t nv, const uint16_t cv)    {      register unsigned short __res;      __asm__ __volatile__ (          LOCK_PREFIX "cmpxchgw %3,(%1)"          : "=a" (__res), "=q" (pv) : "1" (pv), "q" (nv), "0" (cv));      return __res;    }    static __inline__ uint8_t atomic_compare_exchange(volatile uint8_t * pv,         const uint8_t nv, const uint8_t cv)    {      register unsigned char  __res;      __asm__ __volatile__ (          LOCK_PREFIX "cmpxchgb %3,(%1)"          : "=a" (__res), "=q" (pv) : "1" (pv), "q" (nv), "0" (cv));      return __res;    }        typedef void * pvoid;        //function: set *pv to nv    //return: the initial value of *pv    /// 把nv原子地赋值给*pv    static __inline__ pvoid atomic_exchange_pointer(volatile pvoid * pv, const pvoid nv)    {    #ifdef __x86_64__      return (pvoid) atomic_exchange((uint64_t *) pv, (uint64_t) nv);    #else      return (pvoid) atomic_exchange((uint32_t *) pv, (uint32_t) nv);    #endif    }    //function: compare *pv to cv, if equal, set *pv to nv, otherwise do nothing.    //return: the initial value of *pv    /// 比较cv和*pv，如果两者相等则把nv赋值给*pv，并且返回*pv原有数值    /// 否则返回*pv原有数值，不做赋值操作    static __inline__ pvoid atomic_compare_exchange_pointer(volatile pvoid * pv,         const pvoid nv, const pvoid cv)    {    #ifdef __x86_64__      return (pvoid) atomic_compare_exchange((uint64_t *) pv, (uint64_t) nv, (uint64_t)cv);    #else      return (pvoid) atomic_compare_exchange((uint32_t *) pv, (uint32_t) nv, (uint32_t)cv);    #endif    }        #undef LOCK_PREFIX好嘛，又是定义x86下的asm函数，好吧，继续使用gcc的内建__sync的原子操作函数来做实现，修改后的文件如下: