23.03.2007 13:44
OlegON
 
Цитата:
KOT olegon, сделал
Код:
analyze table ffmaprep validate structure cascade
винты реально грузанулись, около часа.
это что такое было?? *09
так в конце-то что написало? error или analyzed?
23.03.2007 13:47
reddevil
 
Цитата:
olegon
Цитата:
reddevil только сейчас обратил внимание -
Цитата:
"analyze table ffmaprep validate structure cascade"
чего этим добиться то хочется?
Дык человек жаловался, что у него оттуда данные не выбираются. Читай выше.
У него это час крутилось значит count(*) - 1фулл тоже не мало времени выполняется.
23.03.2007 13:48
KOT
 
Код:
select count(*) from ffmaprep
сказал что нужны права на просмотр sys.v$statname и sys.v_$sesstatname
и выдал count(*) 5176348
*18
под sys вообще ничего не дает.
Как вообще определиться в каких случаях какой узер нужен? если коротенько.,
23.03.2007 13:56
reddevil
 
ну блин .......
заходишь под sys потом пишешь
alter session set current_schema=SUPERMAG;

выполняешь, потом можеш колбасить select count(*) from ffmaprep и проч.
23.03.2007 13:57
KOT
 
под sys
Код:
select * from v$statname
0 logons cumulative 1
1 logons current 1
2 opened cursors cumulative 1
3 opened cursors current 1
4 user commits 1
5 user rollbacks 1
6 user calls 1
7 recursive calls 1
8 recursive cpu usage 1
9 session logical reads 1
10 session stored procedure space 1
11 CPU used when call started 128
12 CPU used by this session 1
13 session connect time 1
14 process last non-idle time 128
15 session uga memory 1
16 session uga memory max 1
17 messages sent 128
18 messages received 128
19 background timeouts 128
20 session pga memory 1
21 session pga memory max 1
22 enqueue timeouts 4
23 enqueue waits 4
24 enqueue deadlocks 4
25 enqueue requests 4
26 enqueue conversions 4
27 enqueue releases 4
28 global lock sync gets 32
29 global lock async gets 32
30 global lock get time 32
31 global lock sync converts 32
32 global lock async converts 32
33 global lock convert time 32
34 global lock releases 32
35 total file opens 8
36 opens requiring cache replacement 8
37 opens of replaced files 8
38 db block gets 8
39 consistent gets 8
40 physical reads 8
41 db block changes 8
42 consistent changes 8
43 recovery blocks read 8
44 physical writes 8
45 physical writes non checkpoint 8
46 summed dirty queue length 8
47 DBWR checkpoint buffers written 8
48 DBWR transaction table writes 8
49 DBWR undo block writes 8
50 DBWR revisited being-written buffer 8
51 DBWR make free requests 8
52 DBWR free buffers found 8
53 DBWR lru scans 8
54 DBWR summed scan depth 8
55 DBWR buffers scanned 8
56 DBWR checkpoints 8
57 DBWR cross instance writes 40
58 remote instance undo block writes 40
59 remote instance undo header writes 40
60 change write time 8
61 redo synch writes 8
62 redo synch time 8
63 exchange deadlocks 8
64 free buffer requested 8
65 dirty buffers inspected 8
66 pinned buffers inspected 8
67 hot buffers moved to head of LRU 8
68 free buffer inspected 8
69 commit cleanout failures: write disabled 8
70 commit cleanout failures: block lost 8
71 commit cleanout failures: cannot pin 8
72 commit cleanout failures: hot backup in progress 8
73 commit cleanout failures: buffer being written 8
74 commit cleanout failures: callback failure 8
75 commit cleanouts 8
76 commit cleanouts successfully completed 8
77 recovery array reads 8
78 recovery array read time 8
79 CR blocks created 8
80 current blocks converted for CR 8
81 switch current to new buffer 8
82 write clones created in foreground 8
83 write clones created in background 8
84 prefetched blocks 8
85 prefetched blocks aged out before use 8
86 physical reads direct 8
87 physical writes direct 8
88 cold recycle reads 8
89 calls to kcmgcs 128
90 calls to kcmgrs 128
91 calls to kcmgas 128
92 next scns gotten without going to DLM 32
93 Unnecesary process cleanup for SCN batching 32
94 calls to get snapshot scn: kcmgss 32
95 kcmgss waited for batching 32
96 kcmgss read scn without going to DLM 32
97 kcmccs called get current scn 32
98 redo entries 2
99 redo size 2
100 redo buffer allocation retries 2
101 redo wastage 2
102 redo writer latching time 2
103 redo writes 2
104 redo blocks written 2
105 redo write time 2
106 redo log space requests 2
107 redo log space wait time 2
108 redo log switch interrupts 2
109 redo ordering marks 2
110 global cache gets 40
111 global cache get time 40
112 global cache converts 40
113 global cache convert time 40
114 global cache cr blocks received 40
115 global cache cr block receive time 40
116 global cache cr blocks served 40
117 global cache cr block serve time 40
118 global cache cr block send time 40
119 global cache cr block log flushes 40
120 global cache cr block log flush time 40
121 global cache freelist waits 40
122 global cache defers 40
123 global cache convert timeouts 40
124 global cache cr timeouts 40
125 global cache cr requests blocked 40
126 global cache blocks corrupt 40
127 global cache prepare failures 40
128 instance recovery database freeze count 32
129 background checkpoints started 8
130 background checkpoints completed 8
131 serializable aborts 1
132 transaction lock foreground requests 128
133 transaction lock foreground wait time 128
134 transaction lock background gets 128
135 transaction lock background get time 128
136 transaction tables consistent reads - undo records applied 128
137 transaction tables consistent read rollbacks 128
138 data blocks consistent reads - undo records applied 128
139 no work - consistent read gets 128
140 cleanouts only - consistent read gets 128
141 rollbacks only - consistent read gets 128
142 cleanouts and rollbacks - consistent read gets 128
143 rollback changes - undo records applied 128
144 transaction rollbacks 128
145 immediate (CURRENT) block cleanout applications 128
146 immediate (CR) block cleanout applications 128
147 deferred (CURRENT) block cleanout applications 128
148 Commit SCN cached 128
149 Cached Commit SCN referenced 128
150 table scans (short tables) 64
151 table scans (long tables) 64
152 table scans (rowid ranges) 64
153 table scans (cache partitions) 64
154 table scans (direct read) 64
155 table scan rows gotten 64
156 table scan blocks gotten 64
157 table fetch by rowid 64
158 table fetch continued row 64
159 cluster key scans 64
160 cluster key scan block gets 64
161 rows fetched via callback 64
162 leaf node splits 128
163 branch node splits 128
164 native hash arithmetic execute 64
165 native hash arithmetic fail 64
166 index fast full scans (full) 64
167 index fast full scans (rowid ranges) 64
168 index fast full scans (direct read) 64
169 parse time cpu 64
170 parse time elapsed 64
171 parse count (total) 64
172 parse count (hard) 64
173 execute count 64
174 bytes sent via SQL*Net to client 1
175 bytes received via SQL*Net from client 1
176 SQL*Net roundtrips to/from client 1
177 bytes sent via SQL*Net to dblink 1
178 bytes received via SQL*Net from dblink 1
179 SQL*Net roundtrips to/from dblink 1
180 sorts (memory) 64
181 sorts (disk) 64
182 sorts (rows) 64
183 session cursor cache hits 64
184 session cursor cache count 64
185 cursor authentications 128
186 queries parallelized 32
187 DML statements parallelized 32
188 DDL statements parallelized 32
189 DFO trees parallelized 32
190 Parallel operations not downgraded 32
191 Parallel operations downgraded to serial 32
192 Parallel operations downgraded 75 to 99 pct 32
193 Parallel operations downgraded 50 to 75 pct 32
194 Parallel operations downgraded 25 to 50 pct 32
195 Parallel operations downgraded 1 to 25 pct 32
196 PX local messages sent 32
197 PX local messages recv'd 32
198 PX remote messages sent 32
199 PX remote messages recv'd 32
200 buffer is pinned count 72
201 buffer is not pinned count 72
202 no buffer to keep pinned count 72
23.03.2007 13:57
reddevil
 
хотя чето я не понял зачем нам select count(*) from ffmaprep
если только время узнать.
за сколько это выполняется?
23.03.2007 14:04
reddevil
 
мдя)
кот твой тоад умеет план запроса казать?
если да то
Код:
select sum(salesum) from fvmaprep
where saledate between 'даты твоих отчетов';
план и время выполнения сюда
23.03.2007 14:31
KOT
 
1)
Код:
select count(*) from ffmaprep
count(*): 5176348
Цитата:
2) (STRIPE SIZE) Ширина полосы пропускания вроде так по русски. В свойтсвах массива
так и сделано было если о массиве жёстких дисков идет речь.
3)
23.03.2007 14:56
reddevil
 
План и время давай!
23.03.2007 14:58
KOT
 
Люди! Мы уже потерялись в линии обсуждаемой темы, позволю себе маленькое резюме:
1) 1Гб на сервере неприятно, но и не смертельно.
2) правильное выставление
db_block_buffers
db_block_size
shared_pool_size
ускорило отчет до 7-12 минут. УРА!
3)Управляющих файлов достаточно 2
4)
Цитата:
План и время давай!
ка это сделать!?
Часовой пояс GMT +3, время: 20:41.

Форум на базе vBulletin®
Copyright © Jelsoft Enterprises Ltd.
В случае заимствования информации гипертекстовая индексируемая ссылка на Форум обязательна.