Vous êtes sur la page 1sur 25

Limitation du nombre de sessions connectes Dynamic Server

Vous pouvez maintenant limiter le nombre de sessions connectes au serveur. Pour ce faire, dfinissez le
paramtre de configuration LIMITNUMSESSIONS sur le nombre maximum de sessions autorises se
connecter au serveur de base de donnes. Vous pouvez, si vous le souhaitez, spcifier que le serveur
imprime des messages lorsque le nombre de sessions approche le nombre maximum spcifi. Utilisez les
commandes onmode -wm et onmode -wf pour activer/dsactiver ce paramtre de configuration ou
modifier sa valeur

Les valeurs par dfaut des paramtres de configuration Java sont mis jour dans le fichier
onconfig.std

Les paramtres de configuration Java sont dsormais dfinis sur les valeurs par dfaut suivantes dans le
fichier onconfig.std :

JVPJAVAHOME $INFORMIXDIR/extend/krakatoa/jre
JVPHOME $INFORMIXDIR/extend/krakatoa
JVPPROPFILE $INFORMIXDIR/extend/krakatoa/.jvpprops
JVPLOGFILE $INFORMIXDIR/jvp.log
#JDKVERSION 1.5
#JVPARGS -verbose:jni
#JVPCLASSPATH
$INFORMIXDIR/extend/krakatoa/krakatoa_g.jar:$INFORMIXDIR/extend/krakatoa/jdbc_g.jar
JVPCLASSPATH
$INFORMIXDIR/extend/krakatoa/krakatoa.jar:$INFORMIXDIR/extend/krakatoa/jdbc.jar

Reconfiguration du gestionnaire de connexions pendant son excution

Dans les versions prcdentes, vous deviez arrter le gestionnaire de connexions pour modifier sa
configuration, puis le redmarrer pour rendre la nouvelle configuration effective. La nouvelle option de
rechargement (-r) vous permet de reconfigurer le gestionnaire de connexions sans l'arrter , ni le
redmarrer.

Annulation de transactions SQL sur un point de sauvegarde

Vous pouvez dsormais dclarer ou rfrencer des objets de point de sauvegarde dans des instructions
SQL. Un point de sauvegarde identifie un emplacement arbitraire dans les instructions d'une transaction
SQL. Dans sa transaction, le point de sauvegarde ressemble un libell d'instruction sur lequel l'instruction
ROLLBACK SQL peut annuler toute modification apporte la base de donnes gnre par la transaction
entre le point de sauvegarde et l'instruction ROLLBACK. Une application client qui dclare un ou plusieurs
points de sauvegarde dans une transaction peut implmenter une logique de traitement d'erreurs qui
annule uniquement la portion de transaction qui suit le point de sauvegarde indiqu, plutt que d'annuler la
transaction entire si une exception se produit.

Les nouvelles instructions SQL suivantes sont implmentes pour des points de sauvegarde :

L'instruction SAVEPOINT cre un point de sauvegarde dans la transaction SQL en cours.


L'instruction RELEASE SAVEPOINT dtruit un point de sauvegarde spcifi, ainsi que tout autre
point de sauvegarde qui existe entre l'instruction RELEASE et le point de sauvegarde qu'elle
rfrence.
L'instruction ROLLBACK WORK TO SAVEPOINT annule les modifications au schma de base de
donnes ou aux valeurs de ses donnes par les instructions qui suivent le point de sauvegarde,
mais les effets des instructions DDL et DML qui prcdaient le point de sauvegarde persistent.

Cette dition de Dynamic Server prend directement en charge les instructions SQL qui dclarent ou
rfrencent les objets point de sauvegarde dans DB-Access et les UDR crites en langage SPL, C ou Java.
Les mthodes prenant en charge les points de sauvegarde sont galement prises en charge dans les
applications JDBC. Les applications crites en ESQL/C qui utilisent les instructions de point de sauvegarde
SQL sont galement prises en charge dans cette dition de Client SDK.

Modification de la taille de la premire extension d'une table

Lorsque vous crez une table, vous spcifiez la taille d'une premire extension en fonction de la taille
estime de la table. Si la taille de la table est plus grande ou plus petite que son estimation, vous pourriez
souhaiter modifier la taille de la premire extension pour viter de possder un trop grand nombre
d'extensions ou de crer des extensions dont la taille est trop grande. Vous pouvez dsormais modifier la
taille de la premire extension d'une table d'un dbspace. Lorsque vous modifiez la taille de la premire
extension, Dynamic Server enregistre la modification (dans le catalogue systme et la page de partition),
mais n'applique celle-ci que lorsque la table est re-gnre ou qu'une partition ou un fragment est cr.

Cette fonction, y compris la syntaxe de la nouvelle clause MODIFY EXTENT SIZE utilise avec l'instruction
ALTER TABLE

Mise jour dynamique des paramtres de configuration LTXEHWM, LTXHWM et DYNAMIC_LOGS

Vous pouvez dsormais mettre jour de manire dynamique la valeur des paramtres de configuration
LTXEHWM, LTXHWM et DYNAMIC_LOGS l'aide de la commande onmode -wf ou onmode -wm. La
commande onmode -wf modifie la valeur dans le fichier ONCONFIG. La commande onmode -wm modifie
la valeur dans la session en cours.

Nouvelle mise jour automatique des statistiques

Les statistiques sont dsormais automatiquement values pour dterminer si elles doivent tre mises
jour et le sont, le cas chant, chaque nuit. Deux nouvelles tches intgres du planificateur effectuent ce
processus, comme dcrit dans le tableau suivant.

Tableau 3. Tches de mise jour automatique des statistiques

Tche Description

Cette tche value la base de donnes de consignation systme et fournit les


Evaluation des statistiques
commandes de statistiques de mise jour ncessaires en fonction des rgles de
de mise jour automatique
statistiques de mise jour automatique actuelles.

Actualisation automatique
Cette tche excute les commandes de statistiques fournies par la tche Evaluation
des statistiques de mise
des statistiques de mise jour automatique.
jour

Ajout de la prise en charge de la connexion unique

Grce la connexion unique, les utilisateurs s'authentifient une fois et l'authentification est transfre de
manire scuris vers des rseaux de confiance ou non. Les utilisateurs fournissent un ID utilisateur et un
mot de passe valides lorsqu'ils se connectent un ordinateur client. Ils peuvent ensuite accder au serveur
de base de donnes et d'autres services compatibles SSO sans avoir se reconnecter. Auparavant, les
utilisateurs devaient se connecter plusieurs fois.

IDS offre la prise en charge de la connexion unique dans le module Generic Security Services
Communications Support (GSSCSM). GSSCSM fonctionne avec Kerberos, qui est un protocole
d'authentification en rseau. Vous devez dployer un protocole d'authentification Kerberos qui prend en
charge Generic Security Services Application Programming Interface (GSSAPI) avant avant de pouvoir
utiliser SSO avec IDS. Aprs avoir activ SSO, vous bnficiez de la gestion centralise de
l'authentification.

Cette fonction est principalement prsente dans le manuel IBM Informix - Guide de scurit.

Ajout de la prise en charge du chiffrement des donnes l'aide de communications Secure Sockets
Layer (SSL)

Vous pouvez dsormais configurer IDS pour qu'il utilise le protocole SSL, qui chiffre les donnes de
connexions du protocole TCP/IP entre deux points d'un rseau. Le protocole SSL est une alternative au
module de chiffrement CSM spcifique IDS et au mot de passe simple CSM pour les clients CSDK. Vous
devez utiliser SSL pour chiffrer les donnes dans les communications entre les clients IDS et DRDA.

Table 1. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC9

Overview Reference

PN_STAGEBLOB_THRESHOLD configuration
Reserving space for BYTE and TEXT parameter
data in round-robin fragments

If you have tables that contain BYTE or


TEXT columns and that are fragmented
by the round-robin distribution scheme,
you can reserve space for inserting
BYTE and TEXT data by setting the
PN_STAGEBLOB_THRESHOLD
configuration parameter.

This feature is documented in the IBM


Informix Administrator's Reference.
Configuring the server response to low LOW_MEMORY_MGR configuration parameter
memory
onstat -g lmm command: Print low memory
You can configure the actions that the management information
server takes to continue processing
when memory is critically low. You can scheduler lmm enable argument: Specify automatic low
specify the criteria for terminating memory management settings (SQL administration API)
sessions based on idle time, memory
usage, and other factors so that the
targeted application can continue and scheduler lmm disable argument: Stop automatic low
avoid out-of-memory problems. memory management (SQL administration API)
Configuring the low memory response
is useful for embedded applications
that have memory limitations.
Reserving memory for critical activities LOW_MEMORY_RESERVE configuration parameter

You can enable the server to reserve a


specific amount of memory for use
when critical activities (such as rollback
activities) are needed and the server
has limited free memory. If you enable
the new LOW_MEMORY_RESERVE
configuration parameter by setting it to
a specified value in kilobytes, the
critical activities can complete even
when you get out-of-memory errors.
You can also dynamically adjust the
value of the
Table 1. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC9

Overview Reference

LOW_MEMORY_RESERVE
configuration parameter with the
onmode -wm or -wf command.
Support for later versions of IBM GSKIT_VERSION Configuration Parameter
Global Security Kit (GSKit)

Informix 11.50 now supports GSKit


version 7 or later, but ships with GSKit
version 7. If you have a later version of
GSKit software installed on your
computer, you can set the
GSKIT_VERSION configuration
parameter so that the database server
uses the later version.
Handle high severity event alarms Event Alarm IDs

You can plan responses to severity 4 Severity 5 event alarms


and 5 event alarms based on their
explanations and user actions.

The following table lists the new features for Version 11.50.xC8.

Table 2. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC8

Overview Reference

New editions and product names


For more information about the Informix product family,
IBM Informix Dynamic Server editions go to http://www.ibm.com/software/data/informix/.
were withdrawn and new Informix
editions are available. Some products
were also renamed. The publications in
the Informix library pertain to the
following products:
IBM Informix database server,
formerly known as IBM
Informix Dynamic Server (IDS)
IBM OpenAdmin Tool (OAT)
for Informix, formerly known
as OpenAdmin Tool for
Informix Dynamic Server (IDS)
IBM Informix SQL
Warehousing Tool, formerly
known as Informix Warehouse
Feature
Configuring log flow control for RS RSS_FLOW_CONTROL Configuration Parameter
secondary servers

You can limit log activity on the primary


server so that remote standalone (RS)
secondary servers in the cluster can
catch up. This configuration can
improve performance over congested
or intermittent networks. You use the
RSS_FLOW_CONTROL configuration
Table 2. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC8

Overview Reference

parameter to set thresholds that start


and stop flow control.

The following table lists the new features for Version 11.50.xC6.

Table 3. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC6

Overview Reference

The oncmsm Utility


Connection Manager Proxy Support

Connection Manager can now be


configured as a proxy server when
clients connect to Informix data servers
from outside a firewall. You can use
proxy names when you configure high-
availability cluster connections using
the Connection Manager.
Light Scans on Tables BATCHEDREAD_TABLE configuration parameter

You can now enable Informix to onstat -g scn command: Print scan information
perform light scans on compressed
tables, tables with rows that are larger IBM Informix Performance Guide
than a page, and tables with any data,
including VARCHAR, LVARCHAR, and
NVARCHAR data. The server does not
scan pieces of a row (such as smart
large objects) that are stored outside of
the row.

To enable light scans, set the


BATCHEDREAD_TABLE configuration
parameter to 1. You can also use the
IFX_BATCHEDREAD_TABLE
environment option of the SET
ENVIRONMENT statement to change
the value of the
BATCHEDREAD_TABLE configuration
parameter for a session.

To display scan information, use the


new onstat -g scn command. This
command supersedes the onstat -g lsc
command.
Quickly Reverting to Your Source CONVERSION_GUARD configuration parameter
Server After a Failed Upgrade
RESTORE_POINT_DIR configuration parameter
If an upgrade to a new Informix version
or fix pack fails, use the new IBM Informix Migration Guide
onrestorept utility to restore the server
to a consistent, pre-upgrade state. You
can undo changes made during the
upgrade in minutes (and in some
cases, in seconds). Previously, if a fix
pack upgrade failed, you had to restore
the database by using a level-0
Table 3. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC6

Overview Reference

archive.

The new CONVERSION_GUARD and


RESTORE_POINT_DIR configuration
parameters specify information that the
onrestorept utility can use if an
upgrade fails.
Improved Transaction Information onstat -x command: Print database server transaction
information
The onstat -x utility output now
provides an estimated rollback time for
transactions that are rolling back.
Dynamically Start, Stop, or Restart onmode -P: Start, stop, or restart a listen thread
Listen Threads dynamically

You can now dynamically start, stop, or start listen argument: Start a listen thread dynamically
restart (stop and start) an existing listen (SQL administration API)
thread for a SOCTCP or TLITCP
network protocol without interrupting stop listen argument: Stop a listen thread dynamically
existing connections. Use the new (SQL administration API)
onmode -P commands or the new SQL
administration API admin() or task()
functions. restart listen argument: Stop and start a listen thread
dynamically (SQL administration API)
SQL Administration API Portal: SQL administration API portal: Arguments by functional
Arguments by Functional Categories category

You can view admin() and task()


function arguments, listed by category,
and follow the links to information about
the arguments.
Add or Remove BTS Virtual Processors onmode -p: Add or remove virtual processors

You can dynamically add or remove onmode and p arguments: Add or remove virtual
BTS virtual processor by using the processors (SQL administration API)
onmode -p command or the SQL
administration API task() or admin()
function with onmode and p arguments.
View Event Alarms in the Scheduler Events in the ph_alert Table

You can see event alarms in the


ph_alerts table in the sysadmin
database. You can query the ph_alert
table on a local or remote server to
view the recent event alarms for that
server.

The following table lists the new features for Version 11.50.xC5.

Table 4. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC5

Overview Reference

The onclean Utility


Forcing the Database Server to Shut
Table 4. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC5

Overview Reference

Down
The onshutdown script
Use the onclean utility to force the
database server to shut down when
normal shut down with the onmode
utility fails to shut down the server or
when you cannot restart the server.
The onclean utility cleans up shared
memory, semaphores, and stops
database server virtual processes.

Use the onshutdown script to try to shut


down the server normally. If the server
does not shut down after a specified
time, the server is forced to shut down.
onconfig Portal: Configuration onconfig Portal: Configuration parameters by functional
Parameters Listed by Functional category
Categories

You can view configuration parameters


listed in functional categories as they
appear in the onconfig.std file and
follow the links to individual
configuration parameter topics.
SQL Administration API Topics Moved SQL Administration API Overview
to this Guide

The reference topics about all of the


SQL administration API topics are now
in this guide. Previously, these topics
were in the IBM Informix Guide to SQL:
Syntax.
New event alarms with class IDs 73, Event Alarm IDs
74, and 75
Configuring RS Secondary Server DELAY_APPLY Configuration Parameter
Latency for Disaster Recovery STOP_APPLY Configuration Parameter
LOG_STAGING_DIR Configuration Parameter
To aid in disaster recovery scenarios,
you can now configure RS secondary
servers to wait for a specified period of
time before applying logs. Delaying the
application of log files allows you to
recover quickly from erroneous
database modifications by restoring
from the RS secondary server.

The following table lists the new features for Version 11.50.xC4.

Table 5. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC4

Overview Reference

onstat Portal: onstat Utility Commands Sorted


onstat Portal: onstat Commands Listed by by Functional Category
Functional Categories
Table 5. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC4

Overview Reference

You can view onstat commands listed in


functional categories and follow the links to
individual onstat command topics.
Improved Options for Specifying Processor VPCLASS configuration parameter
Affinity

When you specify the processor affinity with the


VPCLASS configuration parameter, you can
now specify individual processors or ranges of
processors. When specifying a range of
processors, you can also specify an
incremental value with the range that indicates
which CPUs in the range should be assigned to
the virtual processors.
Save Disk Space by Compressing Data Compress and Uncompress Operations (SQL
administration API)
You can now use SQL administration API
commands to save disk space by compressing enable compression argument: Enable
row data in a table or in one or more table compression of a table or table fragment (SQL
fragments. You can also use SQL administration API)
administration API commands to consolidate
free space in a table or fragment, return this table or fragment arguments: Compress data
free space to the dbspace, and estimate the and optimize storage (SQL administration API)
amount of space that is saved by compressing
the data.
onstat -g ppd command: Print partition
compression dictionary information
You can display the following types of
information about compression:
onstat -g dsk command: Print the progress of
the currently running compression operation
Active compression dictionaries that
describe how the data is compressed,
syscompdicts_full
with the new onstat -g ppd command
All compression dictionaries, by
querying the new syscompdicts_full oncheck -pt and -pT: Display tblspaces for a
table and syscompdicts view in the Table or Fragment
sysmaster database
Progress of currently running The onlog Utility
compression operations, with the new
onstat-g dsk command IBM Informix Administrator's Guide
Uncompressed contents of
compressed log records, with a new
onlog utility option
Percentage of compressed rows, with
the onstat -pT option

If you are migrating to Version 11.50.xC4 from


Version 11.50.xC1, 11.50.xC2, or 11.50.xC3,
you must run the buildsmi script if you plan to
use the syscompdict_full table. The buildsmi
script, which is in the etc directory in your
installation, drops and recreates the sysmaster
database, which contains the
syscompdict_full table. The buildsmi script
must be run as user informix on UNIX, or as a
member of the Informix-Admin group on
Windows, after ensuring that no connections to
the sysmaster database are made during the
build of the database.
Table 5. What's New in the IBM Informix Administrator's Reference for Version 11.50.xC4

Overview Reference

Enhanced Startup Script Customization The oninit utility


(Windows)

You can now use the oninit -w command to


customize startup scripts and automate startup
on Windows operating systems. The -w flag
forces the command to wait until the server
successfully initializes before returning to the
command prompt. In a high-availability
environment, you can use the oninit -w
command only on primary servers; it is not valid
on secondary servers.
Enable Concurrent I/O to Improve Performance DIRECT_IO Configuration Parameter (UNIX)
on AIX Operating Systems

You can now improve the performance of


cooked files used for dbspace chunks by
setting concurrent I/O, which allows multiple
reads and writes to a file at the same time.
Concurrent I/O provides performance
improvements over direct I/O because it avoids
the serialization of noncompeting reads and
writes from normal file-system locking. Similar
to direct I/O, concurrent IO avoids file system
buffering. Concurrent I/O can be especially
beneficial when you have a file striped across
multiple disks. To enable concurrent I/O on AIX,
set the DIRECT_IO configuration parameter to
2.
Generating a Customized Database Server The genoncfg Utility
Configuration File

Use the new genoncfg utility to generate a


customized Informix configuration file that is
optimized for both anticipated usage and your
host environment.
CHECKALLDOMAINSFORUSER Configuration
New Parameter to Check For Users on All Parameter
Domains

Use the CHECKALLDOMAINSFORUSER


configuration parameter to check all of the
domains for all users.

The following table lists the new features for Version 11.50.xC3.

Table 6. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC3

Overview Reference

onmode -wf, -wm: Dynamically change


Dynamically Updating the LTXEHWM, LTXHWM, certain configuration parameters
and DYNAMIC_LOGS Configuration Parameters.
Table 6. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC3

Overview Reference

You can now dynamically update the value of the


LTXEHWM, LTXHWM, and DYNAMIC_LOGS DYNAMIC_LOGS Configuration Parameter
configuration parameters by using the onmode -
wf or onmode -wm command. The onmode -wf
LTXEHWM Configuration Parameter
command changes the value in the ONCONFIG
file. The onmode -wm command changes the
value for the current session. LTXHWM Configuration Parameter
Using SQL administration API to Dynamically set onconfig memory argument: Temporarily
Update Configuration Parameters change a configuration parameter (SQL
administration API)
You can dynamically set configuration parameters
by using new SQL administration API command set onconfig permanent argument:
arguments. The new function calls emulate Permanently change a configuration
onmode -wf or onmode -wm commands. parameter (SQL administration API)
Use the set onconfig memory argument or the set
onconfig permanent argument to change
configuration parameters in the ONCONFIG file.
Use the set onconfig memory command to
configure parameters only for the current session.
Improved SQL Tracing with the SQL
administration API set sql tracing database argument: Change
database tracing (SQL administration API)
You can use new SQL administration API
arguments to manage SQL tracing by databases, set sql tracing session argument: Control
sessions, and users: set sql tracing database, set tracing for a session (SQL administration
sql tracing session, and set sql tracing user. API)
Previously, you could only trace SQL for all
databases at the server. Now you can control
which databases to include in the SQL trace. You set sql tracing user argument: Control tracing
can also turn tracing on or off for a specific for users (SQL administration API)
session, and specify whether you want to trace
SQL statements run by specific users. set sql tracing argument: Set global SQL
Additionally, you can suspend and resume all tracing (SQL administration API)
tracing at the server, without deallocating any
resources, by using the set sql tracing suspend
and set sql tracing resume arguments.

The following table lists the new features for Version 11.50.xC2.

Table 7. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC2

Overview Reference

BTSCANNER Configuration Parameter


Controlling I/O of B-Tree Indexes with
Compression Levels onmode -C: Control the B-tree scanner

B-tree scanners can now compress indexes by set index compression argument: Change
merging two partially used index pages if the data index page compression (SQL administration
on those pages totals a set level (low, medium, or API)
high). You can specify the index compression
level by modifying the value of the compression
field of the BTSCANNER configuration parameter
option, by running an onmode -C compression
value command, or by running an SQL
Table 7. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC2

Overview Reference

administration API function with a set index


compression command.
Limiting the Number of Sessions That Can LIMITNUMSESSIONS Configuration
Connect to Informix Parameter

You can now limit the number of sessions that


can connect to the server. You do this by setting
the LIMITNUMSESSIONS configuration
parameter to the maximum number of sessions
that you want connected to the database server.
Optionally, you can also specify whether you want
the server to print messages when the number of
sessions approaches a specified maximum
number. You can use onmode -wm and onmode
-wf commands to turn this configuration
parameter on or off or change the value of the
configuration parameter.

The following table lists the new features for Version 11.50.xC1.

Table 8. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC1

Overview Reference

SSL_KEYSTORE_LABEL Configuration
Configuration parameter to use for Secure
Parameter
Sockets Layer (SSL) support

The SSL_KEYSTORE_LABEL configuration


parameter specifies the label of the server digital
certificate used in the keystore database that
stores SSL keys and digital certificates.
Update Data Support on High-Availability UPDATABLE_SECONDARY Configuration
Cluster Secondary Servers Parameter

You can configure secondary servers so that


client applications can send them transactions
that update data. You enable secondary server
updates with the UPDATABLE_SECONDARY
configuration parameter.
Enhanced Connection Management for High- The oncmsm Utility
Availability Clusters

The new Connection Manager dynamically routes


client application connection requests to the most
appropriate server in a high-availability cluster.
Connection Manager connects to each of the
servers in the cluster and gathers statistics about
the type of server, unused workload capacity, and
the current state of the server. From this
information, the Connection Manager redirects
the connection to the appropriate server. In
addition, Connection Manager Arbitrator provides
automatic failover logic for high-availability
clusters. Using a configuration file, you specify
Table 8. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC1

Overview Reference

which secondary server takes over if the primary


server fails.
New utility to encrypt and decrypt password The onpassword Utility
files

The onpassword utility is used to encrypt and


decrypt password files. Password files are used
by Enterprise Replication and Connection
Manager.
Improved onconfig.std file and new default Configuration Parameters
values for configuration parameters

The onconfig.std file is easier to read because


the comments and the parameters are listed
separately and grouped by functional areas. Most
supported configuration parameters are now
included in the file. Deprecated configuration
parameters were removed. Some configuration
parameters that specify sizes now have higher
values. Some configuration parameters that
specify file locations now have more secure
default locations under the $INFORMIXDIR
directory.
Enhanced shared-memory dump file size control DUMPSHMEM Configuration Parameter
(UNIX)
By using the new options for the DUMPSHMEM
configuration parameter and the onstat utility, onstat -o command: Output shared memory
you can control how much memory is written to a contents to a file
dump file. These options exclude the buffer pool
in the resident memory, which can result in a
much smaller file.
Enhanced startup script customization The oninit utility

You can customize startup scripts and automate


startup with the new -w option for the oninit
utility. The -w option forces the server to wait until
it successfully initializes before returning a shell
prompt.
Viewing Data Server client session syssesappinfo
information
onstat -g ses command: Print session-related
A new sysmaster database table and new information
onstat -g ses fields display Data Server client
session ID, session application name, and a
session value.
Alarms for full storage spaces STORAGE_FULL_ALARM Configuration
Parameter
The new STORAGE_FULL_ALARM configuration
parameter sets the threshold level and time
interval for alarms that are raised when storage
spaces or partitions become full.
MSG_DATE Configuration Parameter
Dates for online log messages

The new MSG_DATE configuration parameter


Table 8. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC1

Overview Reference

inserts a date at the beginning of messages


printed to the online log.
HA_ALIAS configuration parameter HA_ALIAS Configuration Parameter

The new HA_ALIAS configuration parameter


represents the name by which the server is
known within a high-availability cluster.
SQL administration API Commands to ha make primary argument: Change the
Configure High-Availability Clusters mode of a secondary server (SQL
administration API)
New commands for the SQL administration API
enable user informix (or members of the DBSA ha rss argument: Create an RS secondary
group who hold Connect privilege on the server (SQL administration API)
sysadmin database and on the new sysha
database) to configure high-availability data ha rss add argument: Add an RS secondary
replication cluster environments remotely. server to a primary server (SQL
administration API)
These SQL function calls emulate the effects of
various onmode -d and onmode -wk ha rss change argument: Change the
command-line options. password of an RS secondary server (SQL
administration API)

ha rss delete argument: Delete an RS


secondary server (SQL administration API)

ha sds clear argument: Stop shared-disk


replication (SQL administration API)

ha sds set argument: Create a shared-disk


primary server (SQL administration API)

ha sds primary argument: Convert an SD


secondary server to a primary server (SQL
administration API)

ha set idxauto argument: Replicate indexes


to secondary servers (SQL administration
API)

ha set ipl argument: Log index builds on the


primary server (SQL administration API)

ha set primary argument: Define an HDR


primary server (SQL administration API)

ha set secondary argument: Define an HDR


secondary server (SQL administration API)

ha set standard argument: Convert an HDR


server into a standard server (SQL
administration API)

ha set timeout argument: Change SD


secondary server timeout (SQL
Table 8. What's New in IBM Informix Dynamic Server Administrator's Reference for Version
11.50.xC1

Overview Reference

administration API)

Tableau 1. Nouveauts dans IBM Informix - Guide de scurit pour Version 11.50.xC2

Prsentation Rfrence

Fonction de fichier de configuration de contrle spcifique au serveur Voir Utilitaire


onshowaudit.
Chaque instance de serveur utilise son propre fichier adtcfg.numserveur si les
paramtres de configuration du contrle sont modifis via l'utilitaire onaudit. La
nouvelle option -n de l'utilitaire onshowaudit lit le fichier de configuration de
contrle spcifique au serveur (adtcfg.numserveur).

Tableau 2. Nouveauts dans IBM Informix - Guide de scurit pour Version 11.50.xC1

Prsentation Rfrence

Prise en charge du chiffrement de donnes l'aide de communications SSL Voir Protocole SSL
(Secure Sockets
Vous pouvez prsent configurer IDS pour utiliser le protocole SSL, qui Layer).
chiffre les donnes des connexions TCP/IP entre deux points d'un rseau.
Le protocole SSL est une alternative au module de support des
communications (CSM) de chiffrement spcifique IDS et au module CSM
avec mot de passe simple pou les clients CSDK. Vous devez utiliser SSL
pour chiffrer les donnes des communications entre des systmes de
dtection d'intrusion et des clients DRDA clients.
Ajout d'une fonctionnalit de prise en charge de connexion unique Voir
Authentification
La connexion unique permet aux utilisateurs de ne s'authentifier qu'une fois SSO.
; cette authentification est poursuivie sur les rseaux scuriss et non
scuriss. Les utilisateurs fournissent un identificateur et un mot de passe
valides lors de la connexion un ordinateur client ; ils peuvent ensuite
accder au serveur de base de donnes et aux autre services compatibles
SSO sans devoir se connecter nouveau. Par le pass, les utilisateurs
devaient se connecter plusieurs fois.

IDS prend en charge la connexion unique dans le module GSSCSM


(Generic Security Services Communications Support Module). GSSCSM
fonctionne avec Kerberos, un protocole d'authentification rseau. Avant
d'utiliser la connexion unique avec IDS, vous devez dployer un protocole
d'authentification Kerberos qui prend en charge l'interface de programme
d'application Generic Security Services (GSSAPI). Une fois la connexion
unique active, la gestion des authentifications est centralise.

Features description with examples

The following features are described in this section:


Forest of trees (FOT) index
Fragment by list and interval
Fragment-level statistics
Storage provisioning
Trusted context
Mapped user
Selective row-level audit (SRLA)
Creating a foreign key constraint with index disabled
Automatic allocation of secondary partition header pages for extending the extent map
Blade manager extensions
New tasks added to Sysadmin database

Forest of trees (FOT) index

A FOT index is an index that is similar to a B-Tree index, but an FOT index has multiple root
nodes and fewer levels. (B-Tree indexes have only one root node and more levels.) You can
think of levels in a FOT index as smaller subtrees (buckets). To retrieve an item from a
subtree, you compute a hash value from the column you chose when creating the index. This
hash value is then mapped to a bucket for storage or for getting a value of the row.

After successfully from converting Informix 11.50.xC7 to 11.70.xC1, you can use the
following command to a create FOT:

Listing 1. SQL syntax for the FOT feature

create index fot_indx on tab_1(col1) hash on (col1) with 1000 buckets;

The command in Listing 1 creates a FOT index on tab_1 table on col1 with 1000 buckets
(subtrees). You can create a FOT on a new table created after conversion or on an existing
table.

Reversion impact: When you try to revert back to an older version, the following message is
displayed, and reversion is cancelled:

Listing 2. Reversion messages

onmode -b 11.50.xC6
Database xdb: Drop FOT index fot_indx before reversion.

Drop the FOT index using the following command:


Listing 3. SQL syntax to drop FOT

drop index fot_indx;

After dropping the FOT index, the reversion will be successful.

System catalog changes: A FOT index creates a new entry in the existing sysindices table.

Fragment by list and interval

List fragmentation fragments data based on a list of discrete values. This type of
fragmentation can help in the logical segregation of data. It is useful when a table has a finite
set of values for the fragment key (or partitioning key) and queries on the table have an
equality predicate on the fragment key.

After successfully converting from Informix 11.50.xC7 to 11.70.xC1, you can use the
following command to create a table with list fragmentation:

Listing 4. SQL syntax for the fragment by list

CREATE TABLE tab_list


(id SERIAL, fname CHAR(32), lname CHAR(32), state CHAR(2), phone CHAR(12))
FRAGMENT BY LIST (state)
PARTITION p0 VALUES ("MA", "NH") IN dbsace0,
PARTITION p1 VALUES ("AZ", "OR") IN dbspace1,
PARTITION p2 VALUES ("KY", "OH") IN dbspace2,
PARTITION p3 VALUES (NULL) IN dbspace3,
PARTITION p4 REMAINDER IN dbspace3;

The command in Listing 4 creates a table called tab_list that fragments data based on column
state using list fragmentation. You can also change the fragmentation scheme of an existing
table to list using the ALTER FRAGMENT command.

Interval fragmentation fragments data based on an interval value. This type of fragmentation
allows the DBA to define a set of initial fragments and an interval value. The DBA does not
have to create all the possible fragments for a table up front. The server automatically creates
fragments based on the specified interval value when rows are inserted that do not fit in the
initial fragments.

After successfully converting from version 11.50.xC7 to 11.70.xC1, you can use the
following command to create table with interval fragmentation:

Listing 5. SQL syntax for the fragment by interval

CREATE TABLE tab_interval (id INTEGER, name CHAR(32), dept CHAR(2), hiredate DATE)
FRAGMENT BY RANGE (id)
INTERVAL (100) STORE IN (dbspace1, dbspace2, dbspace3)
PARTITION p0 VALUES IS NULL IN dbspace0,
PARTITION p1 VALUES < 250 IN dbspace1,
PARTITION p2 VALUES < 500 IN dbspace2;

INSERT INTO tab_interval (id) VALUES (601);

The command in Listing 5 creates a table called tab_interval that fragments data on column id
using interval fragmentation. The table has three initial fragments: p0, p1, and p2. When rows
are inserted with values of id columns >= 500, the server creates interval fragments based on
the specified interval value of 100 in one of the dbspaces specified in the STORE IN clause.
The above insert will result in the creation of an interval fragment with fragment expression
>= 600 and < 700. You can also change the fragmentation scheme of an existing table to
interval using the ALTER FRAGMENT command. Both list and interval fragmentation can be
used to fragment indexes.

Reversion impact: If you have tables fragmented by list or interval, you cannot revert to
11.50.xC7 unless those tables are dropped or the fragmentation scheme of those tables are
modified to make them non-fragmented. When you try to revert back to an older version, the
following message is displayed, and reversion is cancelled:

Listing 6. Reversion messages

onmode -b 11.50.xC6
Database xdb:Table xuser.tab_list:Drop or alter the fragmentation strategy for
Table
before reversion
Database xdb:Table xuser.tab_interval:Drop or alter the fragmentation strategy for
Table
before reversion

After the tables are dropped or altered to make them non-fragmented, the reversion will be
successful.

Listing 7. Drop the tables for list and interval fragmentation

drop tab_list;
drop tab_interval;

System catalog changes:

A new version column is added to sysfragments system catalog table to keep track of
changes to statistics information.
In addition to a row for each table fragment, the sysfragments system catalog table will have
rows for the following information pertaining to list and interval fragmented table or index:
Fragment key (or partitioning key), interval value, and dbspace list for interval fragments.
List fragmentation is represented by 'L,' and interval fragmentation is represented by 'N' in
the strategy column of sysfragments system catalog table.

Fragment-level statistics

This feature helps you to store data distributions per fragment. It allows you to define a data
change threshold, and statistics are refreshed only when the data change threshold is reached.
Run the update statistics command to create or refresh statistics.

After successfully converting from version 11.50.xC7 to 11.70.xC1, ensure you have a valid
smartblob space defined for the SYSSBSPACENAME onconfig parameter, and ensure the space
is allocated. Then use the following command to create fragment level statistics:

Listing 8. SQL syntax for the fragment-level statistics

CREATE TABLE tab_frag_stat


(
s_suppkey integer not null,
s_name char(25),
s_address varchar(40),
s_nationkey integer,
s_phone char(15),
s_acctbal decimal(12,2),
s_comment varchar(101)
)
fragment by expression s_suppkey < 11 in dbspace2,
s_suppkey >= 11 and s_suppkey < 21 in dbspace3,
remainder in dbspace10;

ALTER TABLE tab_frag_stat STATCHANGE AUTO,STATLEVEL FRAGMENT;


UPDATE STATISTICS HIGH FOR TABLE tab_frag_stat;

The STATCHANGE configuration parameter is a new ONCONFIG parameter that is used to define
the threshold. The integer value of the configuration parameter ranges between 0 to 100 or
AUTO; the default value is 10.

The STATLEVEL property is the level used to specify if the distributions must be created at
table level or fragment level. The default value is AUTO. You can also set STATLEVEL TABLE
or STATLEVEL FRAGMENT. When you specify STATLEVEL FRAGMENT, the distributions are
created per fragment.

The command in Listing 8 creates a fragmented table called tab_frag_stat. After you insert
rows and if any of the fragments reach the threshold when update statistics is run, the server
creates the distribution for that particular fragment. The default threshold is 10%. Fragment-
level statistics can be used for new tables created after conversion, or they can be used for
existing tables.

Reversion impact: When you try to revert to an older version, the server takes care of
dropping the catalog changes; there are no reversion messages. Any fragment-level statistics
are dropped, and all tables will use table-level statistics in the older version.
System catalog changes:

A new system catalog table, sysfragdist, is introduced for this feature. This table stores the
distributions per fragment.
New columns ustbuildduration and ustlowts were added to the sysindices table.
New columns statlevel and statchange were added to the systables table.
A new column ustbuildduration was added to sysdistrib table.

New ONCONFIG parameters:

AUTO_STAT_MODE: Has a default value of 1, which enables the automatic update statistics
operation (fragment-level statistics is ON). The value 0 disables automatic update statistics
operation (original server behavior).
STATCHANGE: Has a default value of 10, which defines the percentage of threshold changes.
Integer values for the configuration parameter range from 0 to 100.

Storage provisioning

This feature allows you to automatically allocate a storage space when there is a depletion of
storage space. If you enable this feature, you no longer need DBA intervention to monitor the
space when the application might encounter the problem of depletion of storage space. The
server automatically takes care of adding the space from the storage pool that you specify. A
storage pool can be raw devices, cooked files, or directories. A storage space can be a
dbspace, sbspace, blobspace, temporary dbspace, or temporary sbspace. When there is any
depletion of storage space, existing chunks can be automatically extended, or new chunks can
be automatically added from the storage pool to that particular storage space.

After converting successfully from version 11.50.xC7 to 11.70.xC1, you can use the SQL
administration API commands to implement storage provisioning. For example, to add a
directory called /work/dbspaces/pool1 to the storage pool, specify the following:

Listing 9. Example to add a directory to storage pool

EXECUTE FUNCTION task("storagepool add", "/work/dbspaces/pool1", 0, 0, 1000, 1);

Reversion impact: The storagepool table is automatically dropped during reversion. There
are no reversion messages for this feature.

System database changes:

A new storagepool table is added to sysadmin database.


A new column, is_extendable, is added to the sysmaster:syschunks table.
New columns, create_size and extend_size, are added to sysmaster:sysdbstab table.

New ONCONFIG parameters:

SP_AUTOEXPAND: The default value is 1, which means enable automatic space expansion; 0
means disable automatic space expansion.
SP_WAITTIME: The default value is 30, and the range is 0 to 2147483647.
SP_THRESHOLD: The default value is 0. The server interprets1 to 50 (including decimals) as a
threshold percentage. The server interprets 1000 or above as the number of kilobytes.
Values between 50 to 1000 are invalid.

Trusted context

A trusted context is a database object defining the set of properties for a connection. When the
properties are met, a trusted connection is allowed. The connection is made by a specific user
and made from a trusted client. DBSECADM (Database Security Administrator ) has the
permissions to create, delete, or modify trusted context.

After converting successfully from version11.50.xC7 to 11.70.xC1, you can use the following
command to create trusted context:

Listing 10. SQL syntax for trusted context

CREATE TRUSTED CONTEXT TCX1 BASED UPON CONNECTION USING SYSTEM AUTHID usr3
ATTRIBUTES (ADDRESS '9.99.999.999')
WITH USE FOR usr2 WITHOUT AUTHENTICATION,usr1 WITH AUTHENTICATION ENABLE;

The command in Listing 10 creates a trusted context called TCX1. The address is the IP
address, and when usr3 requests a trusted connection, the request is accepted, because usr3
satisfies all the connection attributes. You can switch the user to usr2 without providing
authentication information, because usr2 is defined as a user of TCX1 without authentication.
However, you cannot switch to usr1 without authentication, because usr1 is defined as a user
of TCX1 with authentication. In order to connect to trusted context, you must have a client,
such as ESQLC, JCC, or JDBC.

Reversion impact: The server takes care of dropping the system catalog changes, and the
reversion goes through without any reversion messages.

System databases changes:

New tables systrustedcontext, systcxattributes, and systcxusers are added to sysusers


database.
New columns succ6, succ7, succ8, fail6, fail7,and fail8 are added to the sysmaster:sysaudit
table.

Mapped user

Currently on UNIX or MS Windows, all the users of the server are also known to the
operating system. Using the mapped user feature, you can externally authenticate a user by
mapping the user to an appropriate user and group privileges, irrespective of whether a user
has an operating systems account. A user who is mapped this way is known as a mapped user.
A DBSA can map a user to an existing operating systems user, or the DBSA can map the user
to a default set of properties.
The authentication can be done using SSO or PAM. For more information on PAM
authentication, take a look at the information on Pluggable Authentication Modules (PAM) in
the IBM Informix Security Guide and in some of the client guides.

After converting successfully from version 11.50.xC7 to 11.70.xC1, you can use the
following command to create a mapped user:

Listing 11. SQL syntax for mapped user

onmode -wf USERMAPPING=BASIC


GRANT ACCESS TO user usr1 WITH PROPERTIES user idsuser;

In order to use this feature, the USERMAPPING configuration parameter can be set to basic or
admin. After authenticating usr1 through PAM, the grant command (used in Listing 11)
maps usr1 to idsuser (the user with the operating systems account), granting the properties of
idsuser to usr1.

Reversion impact: The server drops all of the catalog changes before reversion; the reversion
is successful without any reversion messages.

System database changes: New tables sysusermap, syssurrogategroups, and


syssurrogateusers are added to the sysusers database.

New ONCONFIG parameters:

USERMAPPING: The default value is OFF, which means only users having operating system
accounts can connect to Informix. A value of BASIC means that users without operating
systems account can connect to Informix, but not as privileged users. A value of ADMIN
means that users without operating system accounts can connect to the server as privileged
users.

Selective row-level audit (SRLA)

This feature is an enhancement of existing audit behavior. Using this feature, you can enable
row-level auditing. This improves the performance of the database server. In order to use the
SRLA feature, you must configure the adtcfg file by setting the new ADTROWS configuration
parameter to 1. You must also be a DBSSO.

After converting successfully from Informix 11.50.xC7 to 11.70.xC1, you can use the
following command to create SRLA:

Listing 12. SQL syntax for SRLA

onaudit -c
onaudit -R 1
onaudit -a -u usr1 -e INRW
ALTER table tab_audit ADD AUDIT;

You can check the audit configuration using onaudit -c. The onaudit -R 1 command is
used to enable SRLA. The onaudit -a -u _default -e INRW command creates the user
audit mask usr1, and the granularity is set to Insert Row. You can also set the granularity to
Delete Row (DLRW), Read Row (RDRW), or Update Row (UPRW). The above ALTER command
enables SRLA for the table tab_audit.

You can create a table WITH AUDIT clause or alter table ADD|DROP AUDIT clause.

The command used in Listing 13 displays the generated audit trail.

Listing 13. Utility to check audit trail

onshowaudit

Reversion impact: When you try to revert to an older version, the following message is
displayed, and reversion is cancelled.

Listing 14. Reversion messages

onmode -b 11.50.xC6
Database xdb: Drop the audit property from the table tab_audit before reversion.

Use the command in Listing 15 to drop the SRLA feature:

Listing 15. SQL syntax to drop SRLA

ALTER table tab_audit DROP AUDIT;

After you drop SRLA, the reversion will be successful.

Creating a foreign key constraint with index disabled

New syntax for disabling an index for foreign key constraint was added to use with the ALTER
statement. When creating a foreign key, Informix automatically creates a B-Tree index on the
child table when creating a foreign key. In some situations, the index is not used for any
lookups, there are no updates/deletes from the parent table, and no queries that need this index
on the child table. In this scenario, the B-tree on the child table is simply not needed and
incurs unwanted overhead if the child table is inserted, updated, or deleted.
After converting successfully from version 11.50.xC7 to 11.70.xC1, you can use the
following commands to create a foreign key constraint with index disabled:

Listing 16. SQL syntax for foreign key constraint with index disabled

CREATE table tab_parent (col1 int, col2 int) EXTENT SIZE 16;
CREATE unique index indx_parent_col1 on tab_parent(col1);
ALTER table parent add constraint primary key(col1) constraint cons_parent_col1;
CREATE table tab_child(c1 int, c3 varchar(32)) EXTENT SIZE 16;
CREATE index indx_child_c1 on child(c1);

ALTER table child add constraint


(foreign key(x1) references parent (col1) constraint cons_child_c1 index disabled);

The commands in Listing 16 create a parent table, called tab_ parent, with the primary key
constraint cons_parent_col1. The table has a unique index, indx_parent_col1. The commands
also create child table, called tab_child table with a unique index, indx_chile_c1. The ALTER
command creates the foreign key constraint, cons_child_c1, with the indx_child_c1 index
disabled. During the foreign key enforcement, the server uses the sequential scan, instead of
foreign key index. The foreign key index will be disabled, and the sysobjstate catalog table
will reflect that state. This feature is very useful for data warehouse applications.

Reversion impact: When you try to revert back to an older version, the following message is
displayed, and reversion is cancelled:

Listing 17. Reversion messages

onmode -b 11.50.xC6
Database xdb: Table xuser.child: Drop and recreate this referential constraint
cons_child_x1 before reversion.

To drop the foreign key constraint with index disabled and recreate this referential constraint,
issue the following command:

Listing 18. SQL syntax to drop and recreate the feature

ALTER table child drop constraint cons_child_c1;


ALTER table child add constraint (foreign key(x1) references parent (col1)
constraint cons_child_c1);

After the foreign key constraint with index disabled is dropped and recreated, the reversion
completes successfully.

Automatic allocation of secondary partition header pages for extending the extent map
Prior to 11.70xC1, if the partition page header has a large number of extent entries in it,
conversion would fail, if the server was not able to fit all the extent entries into a single
partition header page. Starting with version 11.70.xC1, if you are converting from a lower
version of Informix, and if a partition header page has a large number of extents entries,
Informix will internally allocate secondary partition header pages (if required) to extend the
extent map, so that conversion will be successful. If more extents are allocated after
converting to 11.70.xC1, and if you try to revert back to an older version, the reversion may
fail, if the server is not able to fit all extent entries into a single partition header page. In order
to successfully revert, you will be required to run defragmenter on such partitions. You can
find more information on Defragment in the IBM Informix 11.70 Information Center (see
Resources").

Blade manager extensions

After conversion, when a UDT, UDR, or INDEX from a database extension (DataBlade
Module) is used, the database extension is automatically registered.

Listing 19. Example where datablade module is automatically registered

CREATE table tab1 (col1 integer, col2 node);

The above command creates a table called tab1, and it will automatically register the node
datablade.

Reversion impact: When you try to revert back to an older version, the following message is
displayed, and reversion is cancelled.

Listing 20. Reversion messages

onmode -b 11.50.xC6
Database xdb: Must drop spl/udr (sysbldsqltextin) before attempting reversion.

There will be more reversion messages apart from the above-mentioned reversion message.

Drop the extensions:

Listing 21. SQL syntax to drop the extensions

drop tab1;
execute function sysbldprepare('Node.*', 'drop');
unprep database name.

After dropping the extensions, the reversion will be successful.


New tasks added to Sysadmin database

New tasks (idle_user_timeout,auto_crsd, check_for_ipa,auto_tune_cpu_vps, and


bad_index_alert) were added to the ph_task table. For the Blade Manager extensions
mentioned above, other new tasks (autoreg exe, autoreg vp, autoreg migrate-console, and
autoreg migrate test) tasks were added to the ph_task table. For the storage provisioning
feature mentioned above, the mon_low_storage task is added to the ph_task table.

Reversion impact: All the new tasks will be dropped by the server before reversion; there are
no reversion messages.

Back to top

Conclusion

This article has provided information that should help you plan how to implement all the new
Version 11.70 features that have potential upgrade and reversion impact. After learning about
this upgrade impact, the system catalog changes, new ONCONFIG parameters, and reversion
impact, you are now ready to take full advantage of all the new features in Version 11.70.

Vous aimerez peut-être aussi