Our db2 log shows some errors recently as following
2005-11-04-10.10.21.593000 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1660 Appid:*LOCAL.DB2.051104021021
database_utilities sqludrsa Probe:0 Database:KOPENDB

Restore terminated.

2005-11-04-10.11.29.031000 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1660 Appid:*LOCAL.DB2.051104021127
buffer_pool_services sqlbinit Probe:103 Database:KOPENDB

The memory for bufferpool and estore cannot be allocated: 0826 0400 .&..


2005-11-04-10.11.29.062000 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1660 Appid:*LOCAL.DB2.051104021127
buffer_pool_services sqlbinit Probe:105 Database:KOPENDB

Database is starting up with tiny bufferpools.
Number of bufferpools coming up: 0100 0000 ....


String Title:sqlbinit PID:924 Node:000
Probe 106. Database is starting up with 1 small buffer pools
each of size 16 pages

2005-11-04-11.21.42.562000 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1072 Appid:none
base_sys_utilities sqleGetAgentFromPool Probe:97
DIA3847C 数据库页标题中发生错误。
(en: database page header occours errors)
ZRC=FFFFFB36

2005-11-04-11.21.42.718000 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1072 Appid:none
base_sys_utilities sqleGetAgent Probe:25

Agent not allocated, sqlcode = -1226

2005-11-04-11.21.42.718001 InstanceB2 Node:000
PID:924(db2syscs.exe) TID:1072 Appid:none
common_communication sqlcctcpconnmgr_child Probe:125
DIA3003E 在 "TCPIP" 协议支持中遇到错误。"sqleGetAgent" 的返回码是 "-1226"。
(en:encountered in "TCPIP" protocol support. Return code from "sqleGetAgent" was "-1226")
sqlcode 1226N
.....and so on

wo do the following ,but fail to resolve it .

1: Increase the "MaxAgents" from 200 to 500 (but I think this database connections no more than 10) ( and "Num_poolAgents" change to 100)
2: run "db2dart dbname" to generate the report . and the report shows no any error
3: reinstall Windows 2000server and DB2 because of finding out some Trojans
4:change the database to another dbserver .
>>>>>>>>>>>

anybody konw how to resolve it ?

Tks

Need the following:

1. Output from db2level

2. Size of each bufferpool defined.

3. Amount of phusical memory on your server.

Pls inspect the page size for buffer pool.
Seemed that your buffer pool is not be allocated since page size issue.

From your messges:
Your database is using 16K pagesize currently. So pls inspect the pagesize for the buffer pool.If the pagesize is not 16K for the buffer pool. Pls set up new buffer pool with 16K pagesize.

DB21085I Instance "DB2" uses
DB2 code release with level "SQL06010"，identifier "01010104" and informational
tokens "db2_v6"、"n990616" and ""

.......
and second about bufferpools
because we change the database to another server(Memory 256M) WE forget to change the buffer size .

___________________

bufferpage size is 4k and we don't know how to change.

__________________________

Original server memory is 1G and database usage 512M

BUFFERPOOL IBMDEFAULTBP SIZE 5000 (4k)

数据库管理程序配置(Database Manager Configuration)



节点类型 = 带有本地客户机和远程客户机的数据库服务器
(Enterprise Server Edition with local and remote clients)


数据库管理程序配置发行版级别 = 0x0900
(Database manager configuration release level)


打开文件的最大总数 (MAXTOTFILOP) = 16000

CPU speed (millisec/instruction) (CPUSPEED) = 6.297920e-007

Max number of concurrently active databases (NUMDB) = 8

Data Links support (DATALINKS) (DATALINKS) = NO

Federated Database System Support (FEDERATED) = NO

Transaction processor monitor name (TP_MON_NAME) =



Default charge-back account (DFT_ACCOUNT_STR) =



Java Development Kit installation path (JDK11_PATH) =



Chapter 5. Practical procedures for high availability 139
Diagnostic error capture level (DIAGLEVEL) = 3

Notify Level (NOTIFYLEVEL) = 2

Diagnostic data directory path (DIAGPATH) =



Default database monitor switches

Buffer pool (DFT_MON_BUFPOOL) = OFF

Lock (DFT_MON_LOCK) = OFF

SORT (DFT_MON_SORT) = OFF

Statement (DFT_MON_STMT) = OFF
TABLE (DFT_MON_TABLE) = OFF
Unit of work (DFT_MON_UOW) = OFF



SYSADM group name (SYSADM_GROUP) =

SYSCTRL group name (SYSCTRL_GROUP) =

SYSMAINT group name (SYSMAINT_GROUP) =



Database manager authentication (AUTHENTICATION) = SERVER

Cataloging allowed without authority (CATALOG_NOAUTH) = NO
Trust all clients (TRUST_ALLCLNTS) = YES

Trusted client authentication (TRUST_CLNTAUTH) = CLIENT



Default database path (DFTDBPATH) = C:



Database monitor heap size (4kb) (MON_HEAP_SZ) = 32

UDF 共享内存集大小(4KB) (UDF_MEM_SZ) = 256
Java Virtual Machine heap size (JAVA_HEAP_SZ) = 512

Audit buffer size (AUDIT_BUF_SZ) = 0
Backup buffer default size(4KB) (BACKBUFSZ) = 1024

Restore buffer default size(4KB) (RESTBUFSZ) = 1024



代理程序堆栈大小 (AGENT_STACK_SZ) = 16

最小的已落实专用内存(4KB) (MIN_PRIV_MEM) = 32

专用内存阈值(4KB) (PRIV_MEM_THRESH) = 1296



排序堆阈值(4KB) (SHEAPTHRES) = 10000



目录高速缓存支持 (DIR_CACHE) = YES



应用程序支持层堆大小(4KB) (ASLHEAPSZ) = 15

最大请求器 I/O 块大小（字节） (RQRIOBLK) = 32767

DOS 请求器 I/O 块大小（字节） (DOS_RQRIOBLK) = 4096

查询堆大小(4KB) (QUERY_HEAP_SZ) = 1000

DRDA 服务堆大小(4KB) (DRDA_HEAP_SZ) = 128



代理程序的优先级 (AGENTPRI) = SYSTEM

现存代理程序的最大数目 (MAXAGENTS) = 500

代理程序池大小 (NUM_POOLAGENTS) = 100

存储池中代理程序的初始数 (NUM_INITAGENTS) = 0

协调代理程序的最大数 (MAX_COORDAGENTS) = MAXAGENTS

并行协调代理程序的最大数 (MAXCAGENTS) = MAX_COORDAGENTS



保持 DARI 进程 (KEEPDARI) = YES

DARI 进程的最大数目 (MAXDARI) = MAX_COORDAGENTS

用 JVM 初始化 DARI 过程 (INITDARI_JVM) = NO

受保护的 DARI 过程的初始数目 (NUM_INITDARIS) = 0



索引重建时间 (INDEXREC) = ACCESS



事务管理程序数据库名 (TM_DATABASE) = 1ST_CONN

事务重新同步间隔（秒） (RESYNC_INTERVAL) = 180



SPM 名 (SPM_NAME) = DBSERVER

SPM 记录大小 (SPM_LOG_FILE_SZ) = 256

SPM 再同步代理程序限制 (SPM_MAX_RESYNC) = 20

SPM 记录路径 (SPM_LOG_PATH) =



NetBIOS 工作站名 (NNAME) =



TCP/IP 服务名 (SVCENAME) = db2cDB2

APPC 事务程序名 (TPNAME) =

IPX/SPX 文件服务器名 (FILESERVER) =

IPX/SPX DB2 服务器对象名 (OBJECTNAME) =

IPX/SPX 套接字号 (IPX_SOCKET) = 879E



Discovery 方式 (DISCOVER) = SEARCH

Discovery 通信协议 (DISCOVER_COMM) = TCPIP

Discover 服务器实例 (DISCOVER_INST) = ENABLE



目录服务类型 (DIR_TYPE) = NONE

目录路径名 (DIR_PATH_NAME) = /.:/subsys/database/

目录目标名 (DIR_OBJ_NAME) =

路由选择信息对象名 (ROUTE_OBJ_NAME) =

缺省客户机通信协议 (DFT_CLIENT_COMM) =

缺省客户机适配器号 (DFT_CLIENT_ADPT) = 0



并行度最大查询级别 (MAX_QUERYDEGREE) = ANY

启用内部分区并行性 (INTRA_PARALLEL) = NO



通信缓冲区数(4KB) (FCM_NUM_BUFFERS) = 1024

FCM 请求块数 (FCM_NUM_RQB) = 512

FCM 连接项数 (FCM_NUM_CONNECT) = (FCM_NUM_RQB * 0.75)

FCM 信息锚数 (FCM_NUM_ANCHORS) = (FCM_NUM_RQB * 0.75)
(SORRY THIS IS chinese output)

Pls check the pagesize for your user tablespace firstly. If it is 16k,Pls set up a new bufferpool with 16K pagesize and assign it to the tablespace.

Did you check whether you have a memory problem ?

Tiny bufferpools are allocated when DB2 cannot allocate the memory needed for the bufferpool.
this usually does not happen during db2start but during activate database. If you don't activate your database explicitely, this will be the first connect.

There should not be a problem (256M of memory vs. 5000x4K~20M for the bufferpools) but I dunno what else you have up an running there. Some java or so can get you out of memory so easy today ...

BTW: for the tiny bufferpools I cant find a ZRC in your diaglog. Was there none or did you omit it here ?

Since the connection itself also fails: check your memory: for each connection a small amount of memory has to be allocated as well (app_ctrl_heap if I remember correctly) and if that is not available you will get an sql code as well (usually that us different from what you get, but who knows).

So what I would do is:
- stop the instance
- start monitor your memory
while monitoring you do:
- start instance (db2start)
- activate database explicitely (db2 activate database) and check db2diag.log for tiny bufferpools
- then try a single connect

--------------------------------
I think when the bufferpool pagesize does not fit with the tablespace specification, it will work a) with one of a smaller pagesize or b) the access to the tablespace is limited. But usually that should not be connected with tiny bufferpools. And also: the catalog is then corrupted and this should not be the case.
the bufferpools are not allocated with 16K pagesize (as try2000 states) but with only 16 pages of 4K(instead of 5000. well, that is tiny).

I try it
______________________

AND why the event moniter get nothing error while database get that errors



_______________________

I think the Memory have no problam .because two server gets the same errors.

only 3 clients
why database create so many agents?
----------
we only create some triggers .does it have some errors?

___________________
trigger s
connect to kopenDB user db2admin using kopen $
drop trigger DB2ADMIN.INSERT $
CREATE TRIGGER DB2ADMIN.INSERT NO CASCADE BEFORE INSERT ON DB2ADMIN.PRODUCT_COSTPRICE REFERENCING NEW AS newrow FOR EACH ROW MODE DB2SQL
BEGIN ATOMIC
set ( newrow.pcp_intax,
newrow.pcp_outtax,
newrow.pcp_cost_price_zb,
newrow.pcp_sale_price,
newrow.pcp_cost_price,
newrow.pcp_classid)=(
select COALESCE(pro_intax,0),
COALESCE(pro_outtax,0),
COALESCE(pro_cost_price,0),
COALESCE(pro_saleprice,0),
COALESCE(pro_cost_price,0),
COALESCE(pro_classid,'')

from product
where pro_komcode= newrow.pcp_pro_komcode);
SET ( newrow.pcp_hold_pur,
newrow.pcp_hold_cus,
newrow.pcp_hold_rep,
newrow.pcp_out_cus,
newrow.pcp_out_rep,
newrow.pcp_balance_stock ,
newrow.pcp_return_stock,
newrow.pcp_in_total,
newrow. pcp_out_total)
= (select COALESCE(sum(st_hold_pur),0),
COALESCE(sum(st_hold_cus),0),
COALESCE(sum(st_hold_rep),0),
COALESCE(sum(st_out_cus),0),
COALESCE(sum(st_out_rep),0),
COALESCE(sum(st_balance_stock),0),
COALESCE(sum(st_return_stock),0),
COALESCE(sum(st_in_total),0),
COALESCE(sum(st_out_total),0)
FROM STOCK
WHERE ST_PRO_KOMCODE = newrow.pcp_pro_komcode and
ST_ORG_ID=newrow.pcp_org_id);
SET NEWROW.PCP_BR_DLHS = CASE WHEN (SELECT COALESCE(BR_TYPE,'000000') FROM BRANCH WHERE BR_ID=NEWROW.PCP_ORG_ID )='000001' THEN
'N'
ELSE (SELECT BR_TAG_dlhs FROM BRANCH WHERE BR_ID=NEWROW.PCP_ORG_ID )
END ;



END
$
CREATE TRIGGER DB2ADMIN.UPDATE_TAX_CLASSID AFTER UPDATE OF PRO_CLASSID, PRO_INTAX, PRO_OUTTAX ON DB2ADMIN.PRODUCT REFERENCING OLD AS oldrow NEW AS newrow FOR EACH ROW MODE DB2SQL when ((select COALESCE(count(*),0) from product_costprice where pcp_pro_komcode=newrow.pro_komcode)>0)
BEGIN ATOMIC

update product_costprice
set
pcp_intax=newrow.pro_intax,
pcp_outtax=newrow.pro_outtax,
pcp_classid=newrow.pro_classid
where pcp_pro_komcode=newrow.pro_komcode;




END
$
DROP TRIGGER DB2ADMIN.TDL_INSERT $
CREATE TRIGGER DB2ADMIN.TDL_INSERT NO CASCADE BEFORE INSERT ON DB2ADMIN.TRADE_DETAIL_LIST REFERENCING NEW AS NEWROW FOR EACH ROW MODE DB2SQL
BEGIN ATOMIC
SET (NEWROW.TDL_AMOUNT_ZB,NEWROW.TDL_OUT_TAX_ZB,NEWROW .TDL_OUT_TOTAL_ZB)=(SELECT NEWROW.TDL_OUT_QTY*PCP_COST_PRICE_ZB,NEWROW.TDL_OU T_QTY*PCP_COST_PRICE_ZB*PCP_OUTTAX,NEWROW.TDL_OUT_ QTY*PCP_COST_PRICE_ZB*(1+PCP_OUTTAX) FROM PRODUCT_COSTPRICE WHERE PCP_PRO_KOMCODE= NEWROW.TDL_PRO_KOMCODE AND PCP_ORG_ID=NEWROW.TDL_ORG_ID);

END
$
------------------

if you set num_pool_agents to 100, that is what you get. The database starts with 0 (num_init_agents) , but wont stop agents processes until it has 100 agents (num_pool_agents)
it is all documented:
http://publib.boulder.ibm.com/infoce...n/r0000145.htm

What event monitor do you have switched on, and why ?
The event monitor does not necessaryly encounter problems because of that, it depends very much on what you are monitoring.

You sure you don't have a memory problem? I mean start monitoring your monitor using operating system tools (whatever you use there), not start an event monitor.
Please check on this.

I would also check again
a) on what system did you database run before
b) what Bufferpool configurations do you have ? I mean there could be more than just the default bufferpool.

If you did have a database usage of 512 M(how was that measured?) this will not fit into 256M of memory. What is your database configuration? is DBHEAP set to something really uge?

re:a) on what system did you database run before
windows 200 server
b) what Bufferpool configurations do you have ? I mean there could be more than just the default bufferpool.
we just use the default bufferpool

we use smartguide to configure