透过pageinspect了解PostgreSQl中的page

  1. 云栖社区>
  2. 博客>
  3. 正文

透过pageinspect了解PostgreSQl中的page

postgresmith 2015-12-31 09:17:01 浏览4955
展开阅读全文
        博客 PostgreSQL的Page分析记录 有过对page的 大体介绍,可以参看之前的blog,下面主要通过插件 pageinspect 向大家进行介绍。 

        在此之前需要了解的名词:
        page,物理文件的单位,默认大小为8K。
        tuple,PG中物理行。
        xid,事务号,执行操作时的顺序id。

        pageinspect里边有三个函数是本文用到的,他们分别是:
        a.get_raw_page,根据参数表明、数据文件类型(main、fsm、vm)以及page位置,将当前表文件中的page内容返回。还有一个函数于此同名,只有两个参数,是将第二个参数省略,直接使用'main'。
        b.page_header,参数是函数get_raw_page的返回值,返回值是将本page结构中的PageHeaderData详细信息
        c.heap_page_items,参数是函数get_raw_page的返回值,返回值是将page内的项指针(ItemIddata)以及HeapTupleHeaderData的详细信息。

        下面通过例子来对page的PageHeaderDataItemIddata和HeapTupleHeaderData进行说明
        a.建表mvcc

[postgres@localhost bin]$ ./psql
psql (9.4.5)
Type "help" for help.
 
postgres=# create table mvcc(id int);
CREATE TABLE
postgres=# insert into mvcc values (1),(2);
INSERT 0 2

         b.PageHeaderData

postgres=# select * from page_header(get_raw_page('mvcc', 'main', 0));
    lsn    | checksum | flags | lower | upper | special | pagesize | version | p
rune_xid
-----------+----------+-------+-------+-------+---------+----------+---------+--
---------
 0/300F4D8 |        0 |     0 |    32 |  8128 |    8192 |     8192 |       4 | 
       0
(1 row)
        这是表mvcc的PageHeaderData信息,现在比对源码对各个字段进行解释:
typedef struct PageHeaderData
{
    /* XXX LSN is member of *any* block, not only page-organized ones */
    PageXLogRecPtr pd_lsn;      /* LSN: next byte after last byte of xlog
                                 * record for last change to this page */
    uint16      pd_checksum;    /* checksum */
    uint16      pd_flags;       /* flag bits, see below */
    LocationIndex pd_lower;     /* offset to start of free space */
    LocationIndex pd_upper;     /* offset to end of free space */
    LocationIndex pd_special;   /* offset to start of special space */
    uint16      pd_pagesize_version;
    TransactionId pd_prune_xid; /* oldest prunable XID, or zero if none */
    ItemIdData  pd_linp[1];     /* beginning of line pointer array */
} PageHeaderData;
        pg_lsn:记录最后一次对page修改的xlog记录id。
        pg_checksum:页面的校验和,主要是通过函数pg_checksum_block函数生成的,0也是有效地,参数为PageHeaderData和BLCKSZ(page's size)。当校验和验证失败,即认为当前页面无效。
        pg_flags:page的flags,具体值为,可以叠加:
#define PD_HAS_FREE_LINES   0x0001      /* are there any unused line pointers? */
#define PD_PAGE_FULL        0x0002      /* not enough free space for new
                                         * tuple? */
#define PD_ALL_VISIBLE      0x0004      /* all tuples on page are visible to
                                         * everyone */
 
#define PD_VALID_FLAG_BITS  0x0007      /* OR of all valid pd_flags bits */
        pg_lower和pg_upper:最后一个项指针的位置和最新的tuple位置。主要进行查找空闲位置,进行插入工作。
        pg_special:page预留的位置,可以存储索引等信息。
        pg_pagesize_version:page大小以及当前版本。page大小可以通过configure进行设置。version的意思是
/*
 * Page layout version number 0 is for pre-7.3 Postgres releases.
 * Releases 7.3 and 7.4 use 1, denoting a new HeapTupleHeader layout.
 * Release 8.0 uses 2; it changed the HeapTupleHeader layout again.
 * Release 8.1 uses 3; it redefined HeapTupleHeader infomask bits.
 * Release 8.3 uses 4; it changed the HeapTupleHeader layout again, and
 *      added the pd_flags field (by stealing some bits from pd_tli),
 *      as well as adding the pd_prune_xid field (which enlarges the header).
 *
 * As of Release 9.3, the checksum version must also be considered when
 * handling pages.
 */
          pg_prune_xid:一般是最后一次删除或者更新的xid。

        pg_linp:项指针。
        c.ItemIddata

postgres=# select lp,lp_off,lp_flags,lp_len from heap_page_items(get_raw_page('mvcc', 'main', 0));
 lp | lp_off | lp_flags | lp_len
----+--------+----------+--------
  1 |   8160 |        1 |     28
  2 |   8128 |        1 |     28
(2 rows)
        这是表mvcc的项指针的信息,一样通过结合源码进行介绍: 
typedef struct ItemIdData
{
    unsigned    lp_off:15,      /* offset to tuple (from start of page) */
                lp_flags:2,     /* state of item pointer, see below */
                lp_len:15;      /* byte length of tuple */
} ItemIdData;
        lp:这是插件自己定义的列,在源码中其实没有,这个是项指针的顺序。 
        lp_off:tuple在page中的位置。 
        lp_flags:tuple的flags,具体为 
#define LP_UNUSED       0       /* unused (should always have lp_len=0) */
#define LP_NORMAL       1       /* used (should always have lp_len>0) */
#define LP_REDIRECT     2       /* HOT redirect (should have lp_len=0) */
#define LP_DEAD         3       /* dead, may or may not have storage */
        lp_len: HeapTupleHeaderData 的长度+Oid的长度(8,因为要数据对齐,所以在这里会比原来预计的多4)。

        d.HeapTupleHeaderData

postgres=# select * from heap_page_items(get_raw_page('mvcc', 'main', 0));
 lp | lp_off | lp_flags | lp_len | t_xmin | t_xmax | t_field3 | t_ctid | t_infom
ask2 | t_infomask | t_hoff | t_bits | t_oid
----+--------+----------+--------+--------+--------+----------+--------+--------
-----+------------+--------+--------+-------
  1 |   8160 |        1 |     28 |   1831 |      0 |        0 | (0,1)  |       
   1 |       2048 |     24 |        |     
  2 |   8128 |        1 |     28 |   1831 |      0 |        0 | (0,2)  |       
   1 |       2048 |     24 |        |     
(2 rows)
         同样结合源码进行介绍: 
typedef struct HeapTupleFields
{
    TransactionId t_xmin;       /* inserting xact ID */
    TransactionId t_xmax;       /* deleting or locking xact ID */
 
    union
    {
        CommandId   t_cid;      /* inserting or deleting command ID, or both */
        TransactionId t_xvac;   /* old-style VACUUM FULL xact ID */
    }           t_field3;
} HeapTupleFields;
 
typedef struct DatumTupleFields
{
 int32      datum_len_;     /* varlena header (do not touch directly!) */
 
 
 int32      datum_typmod;   /* -1, or identifier of a record type */
 
 
 Oid            datum_typeid;   /* composite type OID, or RECORDOID */
 
 
 /*
 * Note: field ordering is chosen with thought that Oid might someday
 * widen to 64 bits.
 */
} DatumTupleFields;
 
struct HeapTupleHeaderData
{
    union
    {
        HeapTupleFields t_heap;
        DatumTupleFields t_datum;
    }           t_choice;
 
    ItemPointerData t_ctid;     /* current TID of this or newer tuple */
 
    /* Fields below here must match MinimalTupleData! */
 
    uint16      t_infomask2;    /* number of attributes + various flags */
 
    uint16      t_infomask;     /* various flag bits, see below */
 
    uint8       t_hoff;         /* sizeof header incl. bitmap, padding */
 
    /* ^ - 23 bytes - ^ */
 
    bits8       t_bits[1];      /* bitmap of NULLs -- VARIABLE LENGTH */
 
    /* MORE DATA FOLLOWS AT END OF STRUCT */
};
        xmin和xmax是插入、删除和更新时的事务ID,插入时会在xmin内写入当前事务ID,当删除时就会在xmax写入当前事务ID。更新是进行删除后再插入。 
        t_cid:这个是指一个事务内的命令ID,每个事务都是从0开始。 
        t_ctid:这个是指物理ID,结构如下: 
typedef struct ItemPointerData
{
    BlockIdData ip_blkid;
    OffsetNumber ip_posid;
}
 
typedef struct BlockIdData
{
    uint16      bi_hi;
    uint16      bi_lo;
} BlockIdData;
        存储的为bi_hi(文件号) << 16 | bi_lo(page号),来获取磁盘顺序,ip_posid是在page的中序号。以此来准确定位数据。 
        t_infomask2:表字段的个数以及一些flags,flags如下: 
#define HEAP_NATTS_MASK         0x07FF  /* 11 bits for number of attributes */
/* bits 0x1800 are available */
#define HEAP_KEYS_UPDATED       0x2000  /* tuple was updated and key cols
                                         * modified, or tuple deleted */
#define HEAP_HOT_UPDATED        0x4000  /* tuple was HOT-updated */
#define HEAP_ONLY_TUPLE         0x8000  /* this is heap-only tuple */
 
#define HEAP2_XACT_MASK         0xE000  /* visibility-related bits */
         t_infomask:tuple的flags,如下:
#define HEAP_HASNULL            0x0001  /* has null attribute(s) */
#define HEAP_HASVARWIDTH        0x0002  /* has variable-width attribute(s) */
#define HEAP_HASEXTERNAL        0x0004  /* has external stored attribute(s) */
#define HEAP_HASOID             0x0008  /* has an object-id field */
#define HEAP_XMAX_KEYSHR_LOCK   0x0010  /* xmax is a key-shared locker */
#define HEAP_COMBOCID           0x0020  /* t_cid is a combo cid */
#define HEAP_XMAX_EXCL_LOCK     0x0040  /* xmax is exclusive locker */
#define HEAP_XMAX_LOCK_ONLY     0x0080  /* xmax, if valid, is only a locker */
 
 /* xmax is a shared locker */
#define HEAP_XMAX_SHR_LOCK  (HEAP_XMAX_EXCL_LOCK | HEAP_XMAX_KEYSHR_LOCK)
 
#define HEAP_LOCK_MASK  (HEAP_XMAX_SHR_LOCK | HEAP_XMAX_EXCL_LOCK | \
                         HEAP_XMAX_KEYSHR_LOCK)
#define HEAP_XMIN_COMMITTED     0x0100  /* t_xmin committed */
#define HEAP_XMIN_INVALID       0x0200  /* t_xmin invalid/aborted */
#define HEAP_XMIN_FROZEN        (HEAP_XMIN_COMMITTED|HEAP_XMIN_INVALID)
#define HEAP_XMAX_COMMITTED     0x0400  /* t_xmax committed */
#define HEAP_XMAX_INVALID       0x0800  /* t_xmax invalid/aborted */
#define HEAP_XMAX_IS_MULTI      0x1000  /* t_xmax is a MultiXactId */
#define HEAP_UPDATED            0x2000  /* this is UPDATEd version of row */
#define HEAP_MOVED_OFF          0x4000  /* moved to another place by pre-9.0
        t_hoff: HeapTupleHeaderData长度,如果有Oid会增加4,但由于受到对齐的影响,会增加8。
        t_bits:具体数据,可以参照 PostgreSQL的基础数据类型分析记录

网友评论

登录后评论
0/500
评论
postgresmith
+ 关注