PG的表膨胀
1 什么是表膨胀
众所周知，PostgreSQL的多版本并发是通过保留变更前的记录来实现的。当数据记录被DML修改，旧版本记录仍保留不变，仅仅需要修改相关记录的xmin、xmax属性，并新增写入变更后的版本记录数据。PG这种直接保留旧版本数据的方式，无需额外的日志写入的消耗，但是一定程度上会导致旧版本数据没有及时清理，就会导致表空间膨胀，影响该表数据的查询效率。
PG数据库为了解决这个问题，主要通过autovacuum和vacuum来进行表膨胀的空间回收利用。默认情况下autovacuum会按照一定的参数设置策略检测并进行一定的清理，但频繁的数据变极大可能导致旧版本数据空间来及进行空间回收时，还是会表空间膨胀的现象。此时是需要数据库运维人员周期性检查是否存在表膨胀，针对表膨胀进行手动vacuum垃圾回收操作。
2 如何判断表膨胀
1、通过 n_dead_tup 来进行计算
select relname,n_live_tup,n_dead_tup,last_autoanalyze,last_autovacuum from pg_stat_all_tables
2、表膨胀TOP 5
SELECT current_database() AS db, schemaname, tablename, reltuples::bigint AS tups, relpages::bigint AS pages, otta, ROUND(CASE WHEN otta=0 OR sml.relpages=0 OR sml.relpages=otta THEN 0.0 ELSE sml.relpages/otta::numeric END,1) AS tbloat, CASE WHEN relpages < otta THEN 0 ELSE relpages::bigint - otta END AS wastedpages, CASE WHEN relpages < otta THEN 0 ELSE bs*(sml.relpages-otta)::bigint END AS wastedbytes, CASE WHEN relpages < otta THEN $$0 bytes$$::text ELSE (bs*(relpages-otta))::bigint || $$ bytes$$ END AS wastedsize, iname, ituples::bigint AS itups, ipages::bigint AS ipages, iotta, ROUND(CASE WHEN iotta=0 OR ipages=0 OR ipages=iotta THEN 0.0 ELSE ipages/iotta::numeric END,1) AS ibloat, CASE WHEN ipages < iotta THEN 0 ELSE ipages::bigint - iotta END AS wastedipages, CASE WHEN ipages < iotta THEN 0 ELSE bs*(ipages-iotta) END AS wastedibytes, CASE WHEN ipages < iotta THEN $$0 bytes$$ ELSE (bs*(ipages-iotta))::bigint || $$ bytes$$ END AS wastedisize, CASE WHEN relpages < otta THEN CASE WHEN ipages < iotta THEN 0 ELSE bs*(ipages-iotta::bigint) END ELSE CASE WHEN ipages < iotta THEN bs*(relpages-otta::bigint) ELSE bs*(relpages-otta::bigint + ipages-iotta::bigint) END END AS totalwastedbytes FROM ( SELECT nn.nspname AS schemaname, cc.relname AS tablename, COALESCE(cc.reltuples,0) AS reltuples, COALESCE(cc.relpages,0) AS relpages, COALESCE(bs,0) AS bs, COALESCE(CEIL((cc.reltuples*((datahdr+ma- (CASE WHEN datahdr%ma=0 THEN ma ELSE datahdr%ma END))+nullhdr2+4))/(bs-20::float)),0) AS otta, COALESCE(c2.relname,$$?$$) AS iname, COALESCE(c2.reltuples,0) AS ituples, COALESCE(c2.relpages,0) AS ipages, COALESCE(CEIL((c2.reltuples*(datahdr-12))/(bs-20::float)),0) AS iotta -- very rough approximation, assumes all cols FROM pg_class cc JOIN pg_namespace nn ON cc.relnamespace = nn.oid AND nn.nspname <> $$information_schema$$ LEFT JOIN ( SELECT ma,bs,foo.nspname,foo.relname, (datawidth+(hdr+ma-(case when hdr%ma=0 THEN ma ELSE hdr%ma END)))::numeric AS datahdr, (maxfracsum*(nullhdr+ma-(case when nullhdr%ma=0 THEN ma ELSE nullhdr%ma END))) AS nullhdr2 FROM ( SELECT ns.nspname, tbl.relname, hdr, ma, bs, SUM((1-coalesce(null_frac,0))*coalesce(avg_width, 2048)) AS datawidth, MAX(coalesce(null_frac,0)) AS maxfracsum, hdr+( SELECT 1+count(*)/8 FROM pg_stats s2 WHERE null_frac<>0 AND s2.schemaname = ns.nspname AND s2.tablename = tbl.relname ) AS nullhdr FROM pg_attribute att JOIN pg_class tbl ON att.attrelid = tbl.oid JOIN pg_namespace ns ON ns.oid = tbl.relnamespace LEFT JOIN pg_stats s ON s.schemaname=ns.nspname AND s.tablename = tbl.relname AND s.inherited=false AND s.attname=att.attname, ( SELECT (SELECT current_setting($$block_size$$)::numeric) AS bs, CASE WHEN SUBSTRING(SPLIT_PART(v, $$ $$, 2) FROM $$#"[0-9]+.[0-9]+#"%$$ for $$#$$) IN ($$8.0$$,$$8.1$$,$$8.2$$) THEN 27 ELSE 23 END AS hdr, CASE WHEN v ~ $$mingw32$$ OR v ~ $$64-bit$$ THEN 8 ELSE 4 END AS ma FROM (SELECT version() AS v) AS foo ) AS constants WHERE att.attnum > 0 AND tbl.relkind=$$r$$ GROUP BY 1,2,3,4,5 ) AS foo ) AS rs ON cc.relname = rs.relname AND nn.nspname = rs.nspname LEFT JOIN pg_index i ON indrelid = cc.oid LEFT JOIN pg_class c2 ON c2.oid = i.indexrelid ) AS sml order by wastedbytes desc limit 5
3、索引膨胀top5
SELECT current_database() AS db, schemaname, tablename, reltuples::bigint AS tups, relpages::bigint AS pages, otta, ROUND(CASE WHEN otta=0 OR sml.relpages=0 OR sml.relpages=otta THEN 0.0 ELSE sml.relpages/otta::numeric END,1) AS tbloat, CASE WHEN relpages < otta THEN 0 ELSE relpages::bigint - otta END AS wastedpages, CASE WHEN relpages < otta THEN 0 ELSE bs*(sml.relpages-otta)::bigint END AS wastedbytes, CASE WHEN relpages < otta THEN $$0 bytes$$::text ELSE (bs*(relpages-otta))::bigint || $$ bytes$$ END AS wastedsize, iname, ituples::bigint AS itups, ipages::bigint AS ipages, iotta, ROUND(CASE WHEN iotta=0 OR ipages=0 OR ipages=iotta THEN 0.0 ELSE ipages/iotta::numeric END,1) AS ibloat, CASE WHEN ipages < iotta THEN 0 ELSE ipages::bigint - iotta END AS wastedipages, CASE WHEN ipages < iotta THEN 0 ELSE bs*(ipages-iotta) END AS wastedibytes, CASE WHEN ipages < iotta THEN $$0 bytes$$ ELSE (bs*(ipages-iotta))::bigint || $$ bytes$$ END AS wastedisize, CASE WHEN relpages < otta THEN CASE WHEN ipages < iotta THEN 0 ELSE bs*(ipages-iotta::bigint) END ELSE CASE WHEN ipages < iotta THEN bs*(relpages-otta::bigint) ELSE bs*(relpages-otta::bigint + ipages-iotta::bigint) END END AS totalwastedbytes FROM ( SELECT nn.nspname AS schemaname, cc.relname AS tablename, COALESCE(cc.reltuples,0) AS reltuples, COALESCE(cc.relpages,0) AS relpages, COALESCE(bs,0) AS bs, COALESCE(CEIL((cc.reltuples*((datahdr+ma- (CASE WHEN datahdr%ma=0 THEN ma ELSE datahdr%ma END))+nullhdr2+4))/(bs-20::float)),0) AS otta, COALESCE(c2.relname,$$?$$) AS iname, COALESCE(c2.reltuples,0) AS ituples, COALESCE(c2.relpages,0) AS ipages, COALESCE(CEIL((c2.reltuples*(datahdr-12))/(bs-20::float)),0) AS iotta -- very rough approximation, assumes all cols FROM pg_class cc JOIN pg_namespace nn ON cc.relnamespace = nn.oid AND nn.nspname <> $$information_schema$$ LEFT JOIN ( SELECT ma,bs,foo.nspname,foo.relname, (datawidth+(hdr+ma-(case when hdr%ma=0 THEN ma ELSE hdr%ma END)))::numeric AS datahdr, (maxfracsum*(nullhdr+ma-(case when nullhdr%ma=0 THEN ma ELSE nullhdr%ma END))) AS nullhdr2 FROM ( SELECT ns.nspname, tbl.relname, hdr, ma, bs, SUM((1-coalesce(null_frac,0))*coalesce(avg_width, 2048)) AS datawidth, MAX(coalesce(null_frac,0)) AS maxfracsum, hdr+( SELECT 1+count(*)/8 FROM pg_stats s2 WHERE null_frac<>0 AND s2.schemaname = ns.nspname AND s2.tablename = tbl.relname ) AS nullhdr FROM pg_attribute att JOIN pg_class tbl ON att.attrelid = tbl.oid JOIN pg_namespace ns ON ns.oid = tbl.relnamespace LEFT JOIN pg_stats s ON s.schemaname=ns.nspname AND s.tablename = tbl.relname AND s.inherited=false AND s.attname=att.attname, ( SELECT (SELECT current_setting($$block_size$$)::numeric) AS bs, CASE WHEN SUBSTRING(SPLIT_PART(v, $$ $$, 2) FROM $$#"[0-9]+.[0-9]+#"%$$ for $$#$$) IN ($$8.0$$,$$8.1$$,$$8.2$$) THEN 27 ELSE 23 END AS hdr, CASE WHEN v ~ $$mingw32$$ OR v ~ $$64-bit$$ THEN 8 ELSE 4 END AS ma FROM (SELECT version() AS v) AS foo ) AS constants WHERE att.attnum > 0 AND tbl.relkind=$$r$$ GROUP BY 1,2,3,4,5 ) AS foo ) AS rs ON cc.relname = rs.relname AND nn.nspname = rs.nspname LEFT JOIN pg_index i ON indrelid = cc.oid LEFT JOIN pg_class c2 ON c2.oid = i.indexrelid ) AS sml order by wastedibytes desc limit 5
3 如何处理表膨胀
1、开启autovacuum，并且根据环境配置设置合理的相关参数
autovacuum_work_mem 、 autovacuum_max_workers 、 maintenance_work_mem ： 保证数据库进行auto vacuum工作进程充足
autovacuum_naptime ：避免该参数设置过大，调小该参数可更加频繁检查和回收膨胀空间
autovacuum_vacuum_scale_factor ： 针对频繁进行批量DML操作的表，可以调小该参数，保证DML操作后可尽快进行回收膨胀空间
2、使用pg_reorg或者vacuum full手动回收膨胀的空间
