eygle.com   eygle.com
eygle.com  
 
留言簿 - Oracle Life - Powered by Eygle.com
eygle.com 我要留言
北京恩墨科技有限公司成立
昵称
内容 页: < 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 > >> - 403
# 45917
小骨




To: eygle
  前辈,您好~
最近在学oracle b-tree index,在文档上说,b-tree的叶子节点是双向链表,但我dump出来在trc文件里面没有发现rlink和llink
rlink:指向直接后继
llink:指向直接前驱
不知道这两个指针(rowid)被置于哪里了?
谢谢~
From: 小骨
2012.08.16 07:57
To: 小骨
  是 kdxlenxt / kdxleprv 分别代表下一个叶节点何前一个叶节点。没有你说得 rlink / llink.
From: eygle
2012.08.20 07:42

版主选项: 回复 编辑
# 45916
vast




To: eygle
  eygle老师 我的数据库做了一次检查发现buffer hit 的命中率过低 在70%左右,请问产生的具体原因和解决办法。
From: vast
2012.08.15 04:26
To: vast
  增大Buffer Cache设置,可能因为你的全表扫描太多了。
From: eygle
2012.08.19 05:04

版主选项: 回复 编辑
# 45915
kamus1989


来自: 西安


To: eygle
  2012-08-14 04:46:57.303 [21] >TRACE: clssgmReconfigThread:completed for reconfig(2), with status(1)
2012-08-14 04:46:57.763 [17] >TRACE: clssgmCommonAddMember: clsomon joined (1/0x1000000/#CSS_CLSSOMON)
From: kamus1989
2012.08.14 19:16

版主选项: 回复 编辑
# 45914
kamus1989


来自: 西安


To: eygle
  2012-08-14 04:46:57.296 [1] >USER:NMEVENT_SUSPEND [00][00][00][00]
2012-08-14 04:46:57.297 [21] >TRACE: clssgmReconfigThread:started for reconfig (2)
2012-08-14 04:46:57.298 [21] >USER:NMEVENT_RECONFIG [00][00][00][06]
2012-08-14 04:46:57.298 [21] >TRACE: clssgmEstablishConnections: 2 nodes in cluster incarn 2
2012-08-14 04:46:57.299 [17] >TRACE: clssgmInitialRecv: (6000000000ab65a0) accepted a new connection from node 2 born at 1 active (2, 2), vers (1
0,3,1,2)
2012-08-14 04:46:57.299 [17] >TRACE: clssgmInitialRecv: conns done (2/2)
2012-08-14 04:46:57.300 [21] >TRACE: clssgmEstablishMasterNode: MASTER for 2 is node(2) birth(1)
2012-08-14 04:46:57.300 [21] >TRACE: clssgmChangeMasterNode: requeued 0 RPCs
2012-08-14 04:46:57.302 [17] >TRACE: clssgmHandleDBDone(): src/dest (2/65535) size(72) incarn 2
CLSS-3000: reconfiguration successful, incarnation 2 with 2 nodes
CLSS-3001: local node number 1, master node number 2
From: kamus1989
2012.08.14 19:15

版主选项: 回复 编辑
# 45913
kamus1989




To: eygle
  eygle老师,您好!
 我们之前做了一次数据库搬迁,直接将源库的存储挂在新的机器上。可是crs_stat -t 中两个节点的实例status都是offline。通过srvctl启动实例,status会在unknown和offline之间切换。存储挂完后,将CRS中的默认spfile的位置指定到存储上spfile的位置。
 新机器的操作系统补丁包比源机器的新,数据库和crs的版本都是一样的。
OS版本:HP-UX B.11.31
DB和CRS版本:10.2.0.4.0
alerts60bild1.log:
2012-08-14 04:46:57.031
[cssd(8993)]CRS-1605:CSSD voting file is online: /dev/lockvgbill/rrac_voting_1. Details in /oracle/crs/log/s60bild1/cssd/ocssd.log.
ocssd.log:
2012-08-14 04:46:57.205 [13] >TRACE: clssnmUpdateNodeState: node 2, state (4/3) unique (1344890251/1344890251) prevConuni(0) birth (0/1) (old/new)
2012-08-14 04:46:57.205 [13] >USER:clssnmHandleUpdate: SYNC(2) from node(2) completed
2012-08-14 04:46:57.205 [13] >USER:clssnmHandleUpdate: NODE 1 (s60bild1) IS ACTIVE MEMBER OF CLUSTER
2012-08-14 04:46:57.205 [13] >USER:clssnmHandleUpdate: NODE 2 (s70bild2) IS ACTIVE MEMBER OF CLUSTER
2012-08-14 04:46:57.205 [13] >TRACE: clssnmHandleUpdate: diskTimeout set to (597000)ms
From: kamus1989
2012.08.14 19:14

版主选项: 回复 编辑
# 45912
xiaofagn




To: eygle
  *** ACTION NAME0000312 STARTED62) 2012-08-14 04:14:15.415
ksfqfret:ctx=0xacccab38 filename=/crgw/backup1/ora10g_1/ora10g/data/20120814_2/data.ORA10G_gfnijgpi_1_1_791266098.live.2 blksiz=8192
 startblk=1
ksfqqrdffset=1 nblocks=127 filsiz=434
ksfqrd:ctx=0xacccab38 nblocks=127 wait=1
ksfqrd:buf=0xacf57e00 offset=1 blksread=127
*** ACTION NAME0000313 STARTED62) 2012-08-14 04:14:15.783
ksfqfret:ctx=0xacccab38 filename=/crgw/backup1/ora10g_1/ora10g/data/20120814_2/data.ORA10G_ggnijgu3_1_1_791266243.live.2 blksiz=8192
 startblk=1
ksfqqrdffset=1 nblocks=127 filsiz=4348
ksfqrd:ctx=0xacccab38 nblocks=127 wait=1
ksfqrd:buf=0xacf57e00 offset=1 blksread=127
*** ACTION NAME0000314 STARTED62) 2012-08-14 04:14:18.685
ksfqfret:ctx=0xacccab38 filename=/crgw/backup1/ora10g_1/ora10g/data/20120808_3/arch.ORA10G_efni3kpk_1_1_790745908_live.1 blksiz=512
startblk=1
ksfqrd:ctx=0xacccab38 nblocks=127 wait=1
ksfqrd:buf=0xacf57e00 offset=1 blksread=2048
From: xiaofagn
2012.08.14 18:08

版主选项: 回复 编辑
# 45911
flybuffer




To: eygle
  请问dba手记3 有全部源码的下载吗 你的网站上只有部分源码 书上也没见下载地址
From: flybuffer
2012.08.08 02:57
To: flybuffer
  没有,如果看到哪些代码过长,在这里留下页码,我帮助摘录出来,不便之初,敬请谅解。
From: eygle
2012.08.08 19:48

版主选项: 回复 编辑
# 45910
shadow


来自: 深圳


To: eygle
  
From: shadow
2012.08.07 01:40
To: shadow
  这个问题是很复杂的,不是一两句能说清楚的。
From: eygle
2012.08.07 08:14

版主选项: 回复 编辑
# 45909
gao




To:
  盖老师您好,在您的深入解析oracle一书的59页“而只是记录控制文件检查点scn并且根据检查点的写出增进rba信息” 这里的记录控制文件检查点scn是指什么?是指当ckpt进行轻量级更新时将当前的scn更新到controlfilecheckpoint at scn吗?可我在10g的版本中测试的时候发现controlfilecheckpoint at scn的值并没有改变。
From: gao
2012.08.01 05:43
To: gao
  参考62页信息,如果CCSCN未改变,那就是增量检查点也没有发生,或者数据库中没有产生Dirty Buffer,没有相关信息写出。
From: eygle
2012.08.05 04:27

版主选项: 回复 编辑
# 45908
puck




To: eygle
  It is very honored for me to read your articles . I have some doubts about the data world. Could you tell me how do you think about HADOOP,HBASE,NOSQL and ORACLE in large data era.Can the traditional rdbms live in the coming years.THANKS.
From: puck
2012.07.31 18:24
To: puck
  RDBMS不会衰亡,HBASE,NOSQL会各占一份细分市场,乱花渐欲迷人眼!
From: eygle
2012.08.02 18:22

版主选项: 回复 编辑

页: < 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 > >> - 403
我要留言
Copyright © 2003~2012 eygle.com All Rights Reserved.
Powered by: www.eygle.com