CQL Join on the coherence of Cache with Composite key

Similar Questions

• How to load a multiple column table in the coherence of caches?

  How to load a multiple column table in the coherence of caches?
  I want to load a multi-column (about 20 columns) table in cache coherence. How to change the following code (spatially how to change the SQL SELECT statement)?
  Is the following select statement enough: Select the key, the value of EMPLOYEES ?
  

  public static void bulkLoad(NamedCache cache, Connection conn)
      {
      Statement s;
      ResultSet rs;
      
      try
          {
          s = conn.createStatement();
          rs = s.executeQuery("select key, value from table");
          while (rs.next())
              {
              Integer key   = new Integer(rs.getInt(1));
              String  value = rs.getString(2);
              cache.put(key, value);
              }
          ...
          }
      catch (SQLException e)
          {...}
      }

  First of all, you need a class to hold your 20 fields:

  public class Data {
  
      private String field1;
      private String field2;
      private String field3;
      private String field4;
      private String field5;
      private String field6;
      private String field7;
      private String field8;
      private String field9;
      private String field10;
      private String field11;
      private String field12;
      private String field13;
      private String field14;
      private String field15;
      private String field16;
      private String field17;
      private String field18;
      private String field19;
      private String field20;
  
      public Data() {
      }
  
      public String getField1() {
          return field1;
      }
  
      public void setField1(String field1) {
          this.field1 = field1;
      }
  
      public String getField2() {
          return field2;
      }
  
      public void setField2(String field2) {
          this.field2 = field2;
      }
  
      public String getField3() {
          return field3;
      }
  
      public void setField3(String field3) {
          this.field3 = field3;
      }
  
      public String getField4() {
          return field4;
      }
  
      public void setField4(String field4) {
          this.field4 = field4;
      }
  
      public String getField5() {
          return field5;
      }
  
      public void setField5(String field5) {
          this.field5 = field5;
      }
  
      public String getField6() {
          return field6;
      }
  
      public void setField6(String field6) {
          this.field6 = field6;
      }
  
      public String getField7() {
          return field7;
      }
  
      public void setField7(String field7) {
          this.field7 = field7;
      }
  
      public String getField8() {
          return field8;
      }
  
      public void setField8(String field8) {
          this.field8 = field8;
      }
  
      public String getField9() {
          return field9;
      }
  
      public void setField9(String field9) {
          this.field9 = field9;
      }
  
      public String getField10() {
          return field10;
      }
  
      public void setField10(String field10) {
          this.field10 = field10;
      }
  
      public String getField11() {
          return field11;
      }
  
      public void setField11(String field11) {
          this.field11 = field11;
      }
  
      public String getField12() {
          return field12;
      }
  
      public void setField12(String field12) {
          this.field12 = field12;
      }
  
      public String getField13() {
          return field13;
      }
  
      public void setField13(String field13) {
          this.field13 = field13;
      }
  
      public String getField14() {
          return field14;
      }
  
      public void setField14(String field14) {
          this.field14 = field14;
      }
  
      public String getField15() {
          return field15;
      }
  
      public void setField15(String field15) {
          this.field15 = field15;
      }
  
      public String getField16() {
          return field16;
      }
  
      public void setField16(String field16) {
          this.field16 = field16;
      }
  
      public String getField17() {
          return field17;
      }
  
      public void setField17(String field17) {
          this.field17 = field17;
      }
  
      public String getField18() {
          return field18;
      }
  
      public void setField18(String field18) {
          this.field18 = field18;
      }
  
      public String getField19() {
          return field19;
      }
  
      public void setField19(String field19) {
          this.field19 = field19;
      }
  
      public String getField20() {
          return field20;
      }
  
      public void setField20(String field20) {
          this.field20 = field20;
      }
  }
  

  Then you can use it to store data in you original code

  public static void bulkLoad(NamedCache cache, Connection conn)
      {
      Statement s;
      ResultSet rs;
  
      try
          {
          s = conn.createStatement();
          String sql = "select key, value, value2, " +
                  "value3, value4, value5, value6, " +
                  "value7, value8, value9, value10 " +
                  "value11, value12, value13, value14, " +
                  "value15, value16, value17, value18, " +
                  "value19, value120 from table";
  
          rs = s.executeQuery(sql);
          while (rs.next())
              {
              Integer key   = new Integer(rs.getInt(1));
              Data data = new Data();
              data.setField1(rs.getString(2));
              data.setField2(rs.getString(3));
              data.setField3(rs.getString(4));
              data.setField4(rs.getString(5));
              data.setField5(rs.getString(6));
              data.setField6(rs.getString(7));
              data.setField7(rs.getString(8));
              data.setField8(rs.getString(9));
              data.setField9(rs.getString(10));
              data.setField10(rs.getString(11));
              data.setField11(rs.getString(12));
              data.setField12(rs.getString(13));
              data.setField13(rs.getString(14));
              data.setField14(rs.getString(15));
              data.setField15(rs.getString(16));
              data.setField16(rs.getString(17));
              data.setField17(rs.getString(18));
              data.setField18(rs.getString(19));
              data.setField19(rs.getString(20));
              data.setField20(rs.getString(21));
              cache.put(key, data);
              }
          ...
          }
      catch (SQLException e)
          {...}
      }    
  

  Of course your data object would need the appropriate field names and you will have no channels for all types. It must also implement equals and hashCode and ideally PortableObject.

  JK

• is it better to use the rowid function or a composite key

  I have a clue about 2 fields so that I can make my form based on these 2 fields
  But now since I spend my app in 4.2 that I intend to do everything depended of rowid function
  
  Is this the right approach
  I currently facing problems as 1422 extract extract several lines even if the primary key is defined as key composite column 2

  Your primary key must never rest on 'natural touches', that is to say the actual data in your table. The fact that you are talking about a composite key means that you think that do. You can create an index that is UNIQUE on these types of fields to maintain the integrity of the data, but they do not have the pharmacokinetics of the table.

  My primary keys are always a NUMERIC field, populated by a combination of sequence/relaxation. Because the data in the ID field have no connection with the data in the table - there is never any need/reason to change. As has been mentioned above, ROWID changes rarely, but it is possible for them. I use ROWID in the Apex and when I have forms to keep tiny (maybe 5-10 lines) lookup tables that are used to populate LOVs create a primary key is excessive.

• Reactivation of the streets & trips 2008 with product key won't work.

  My computer hard drive crashed.  I bought a new computer.  I use to reinstall Activation Assistant streets & trips 2008 but he said that the product key has been used too many times.
  For customer technical support phone call said that a technician on this site to reactivate it for me?

  Hi Sylvia,.

  You can follow the steps in the article mentioned below and check:

   

  Solve problems with activation for Microsoft games problems and mapping programs

  http://support.Microsoft.com/kb/960480

  I would also recommend posting your query in the Support of Microsoft Streets & Trips for better assistance.

   

  Microsoft Streets & Trips Support

  http://www.Microsoft.com/streets/en-us/support-and-training.aspx

  Support for MapPoint, streets & trips and Highway

  http://support.Microsoft.com/ph/1058

  Hope this information helps.

• How to re - enter the product originally supplied with Windows key

  I used my computer for over 2 years & nothing happened. But for about 1 month ago, my PC has its hard drive broke. So I brought in a repair shop. But now it always wonder about a bit as the key active product or u are a victim of... I discovered that I have to re - enter the product key, or buy a new one. But it will cost a lot of money. What can I do now?

  You're the edition which is now installed corresponds with which is preinstalled on your computer? Its possible to maybe installed repair facility the wrong version.

  Click Start, right click on computer

  Click on properties

  See editing

  Compare with editing on the COA sticker on your machine:

  COA certificate of authenticity:

  http://www.Microsoft.com/howtotell/content.aspx?PG=COA

  What is the certificate of authenticity for Windows?

  http://Windows.Microsoft.com/en-us/Windows7/what-is-the-Windows-certificate-of-authenticity

  If they do not match, you need to install the correct edition. If this is the right edition, enter the product key on the COA sticker.

  Press Windows key + R

  Type: slui.exe 3

  Press enter

  Enter the product key

  Click next to activate via Internet

• Integrate Hibernate L2 cache with consistency in the composite key of reference

  Hi all:
  I have a problem when I activate the Hibernate l2 cache integrated with consistency. When the entity class using this string or unique integer primary key, the L2 cache has worked well. But when I use the entity with composite key class, I had the problem.
  I wrote a script to test junit to test the session.get method, when I run the test cases for the first time, the miss of the entity to load the object from the cache, Hibernate will trigger a DB query so it box as an object, then put it in the cache. But when I run the test case, again, I found the Hibernate trigger a query once again, which was strange... (it should not have happened, I should got the object from the cache), then, I used the console cache coherence to list the data in the cache and finding himself under the information:
  ###########################################
  (Hi) map: list
  model. SiteUserInfo #model. SiteUserID@323819 = Item {version = null, freshTimestamp = 1333943690484
  model. SiteUserInfo #model. SiteUserID@323819 = Item {version = null, freshTimestamp = 1333943600419
  ###########################################
  There are two keys with the same value cached... (I put already implemented method hashcode and equals to SiteUserID object). someone there has the same problem? Or L2 Hibernate with composite key can integrate with consistency?
  
  Rip off code SiteUserID
  #####################################
  @Override
  public boolean equals (Object obj) {}
  If (obj == this) {}
  Returns true;
  }
  If (!) () obj instanceof SiteUserID)) {}
  Returns false;
  }
  SiteUserID id = (SiteUserID) obj;
  return new EqualsBuilder () .append (sid, id.getSid ()) .append (name, id.getName ()) .isEquals ();
  }
  
  @Override
  public int hashCode() {}
  return new HashCodeBuilder () .append (sid) .append (name) .toHashCode ();
  }
  ###################################

  Hi John,.

  I think there may be another problem.

  Please reference:
  http://blackbeanbag.NET/WP/2010/06/06/coherence-key-HOWTO/
  For entry cache, partitioned cache is placed in backup card in binary form.
  The main objects of the same data must always be serialized in binary form even.

  Unfortunately, use Hibernate CacheKey is as key cache L2.
  And CacheKey is somewhat complex, some related attributes use TreeMap as their type.
  "Java gives no guarantee on the order in a hash table and serialization is dependent on the command."
  (reference http://stackoverflow.com/questions/9337126/how-can-oracle-coherence-get-fail-with-retrieved-key-object)

  So... you may need to modify the source code of some Hibernate, ex: replace the FastHashMap in TreeMap...
  I tried, it looks like work. I'll send you have modified source codes.

  Best regards
  Leon

• not the substitution of the coherence-cache - config.xml, but showing the error...

  Hello
  
  I created a file called file tangosol-coherence - override.xml in the specified path which is "C:\Program Files\Oracle\Coherence for.NET"
  
  and the coherence.jar file is located in the path that is "C:\Program Files\Oracle\Coherence for." NET\lib ".
  
  but it is not the substitution of the XML... with the default file...
  
  
  C:\Program Files\Oracle\Coherence pour.NET\examples\ContactCache.Java > C:\Progra «»
  m Files\Java\jdk1.6.0_11\bin\java"- Server showversion-Xms128m-Xmx128m - Dtang
  STATUS ONLINE. Coherence.TTL = 0 - Dtangosol.coherence.cacheconfig = contact-cache - config.xml.
  CP 'config; lib\custom - types.jar; C:\Program Files\Oracle\Coherence for .NET\lib\c
  OHERENCE.jar' com.tangosol.net.DefaultCacheServer
  Java version "1.6.0_11.
  Java (TM) SE Runtime Environment (build 1.6.0_11 - b03)
  Java hotspot Server VM (build 11, 0 - b16, mixed mode)
  
  2011-06-20 15:06:57.607/2.366 Oracle coherence 3.7.0.0 < Info > (thread = hand, member)
  ER = n/a): responsible operational configuration of jar:file:/C:/Program%20Files/Ora «»
  cle/Coherence%20for%20.NET/lib/coherence.jar!/tangosol-coherence.xml ".
  2011-06-20 15:06:57.989/2.748 Oracle coherence 3.7.0.0 < Info > (thread = hand, member)
  ER = n/a): operational substitutions of jar:file:/C:/Program%20Files/Oracle/ load «»
  Coherence%20For%20.NET/lib/coherence.jar!/tangosol-coherence-override-dev.xml ".
  2011-06-20 15:06:58.049/2.808 Oracle coherence 3.7.0.0 < D5 > (thread = main Member,
  (= n/a): configuration optional override ' / tangosol-coherence - override.xml ' is not
  specified
  2011-06-20 15:06:58.062/2.821 Oracle coherence 3.7.0.0 < D5 > (thread = main Member,
  (= n/a): configuration optional override "/ custom - mbeans.xml ' is not specified
  
  Oracle Version 3.7.0.0 Build 23397 consistency
  Grid edition: development Mode
  
  Thanks in advance

  The classpath is not the same as the path. It is similar, but a classpath is a set of directories or jar Java files use to search for resources and executable code.

  Assuming you are using the sample file of Start-cache - server.cmd which is supplied with the .net examples to start your server you can put the file tangosol-coherence - override.xml in the C:\Program Files\Oracle\Coherence pour.NET\examples\ContactCache.Java\config directory because it's on the class path when using this script.

  JK

• Is possible to change the type of node WINS with cmd command on a windows operating system that is not joined to the domain?

  Is possible to change the type of node WINS with cmd command on a windows operating system that is not joined to the domain?

  Hello

  The question you posted would be better suited to the TechNet community. Please visit the link below to find a community that will provide the support you want.

   
  http://social.technet.Microsoft.com/forums/en-us/itprovistanetworking/threads
   

  Hope this information is useful.

• installation stops with a "cannot detect the last level cache.

  ESXi 5.0 installation process stops with a message "cannot detect the last level cache.

  VMware ESXi 5.0.0 [Releasebuild-469512 x86_64]
  Cannot detect the last level cache
  CR0 = 80010039 cr2 = 0 x 0 = 0x3e7000 cr4 = 0 x 2038 cr3
  PCPU0:2048 / bootstrap
  PCPU 0: S
  Beginning of the code: uptime VMK 0 x 418013200000: 0:00:00:00.890
  6 lines of < no symbol > errors
  Basic FS = 0 x 0 gs = 0 x 418040000000 kg = 0 x 0
  No space on disk for the dump data.
  Debugger waiting...

  switching LSI based SAS controller to port sata3 onboard has not changed a thing

  Oh, here the specifications:
  Processor Intel i5 2500
  16 GB OF RAM
  ASUS P8H67-M Board

  any ideas?

  Hi all

  I had the same problem after turning on in BIOS "Max CPUID Value Limit" to activate option.

  Disabling above option help me to install vmware esxi 5.0

  Anatoliy

• What is the scope of queries cached with cachedwithin on CF10? Server or application?

  Hello

  We have recently migrated to CF8 to CF10 and noticed an unexpected behavior on our queries cached.

  We are a news site and have two applications on our server (CF 10, 5.8 64-bit Centos), one for the front end for the readers and the other for the administrators of the site. We have made extensive use of queries cached, always with cachedwithin and always on CFCS. When a user reads an article, request that retrieve the article is put in cache for 15 minutes, and if an administrator makes a change on this article, after the update, we call exactly the same read request with a cachedwithin (0,0,0,0) to clear the data cache.

  It worked well until upgrading to CF 10, when we realized that calls for cleaning of the application (s) were not working. After many tests we noticed that if we call the inner query cache cleaning the scope of readers, then it is properly cleared the cache for the readers. The same exact query called administrators field only clears the cache for administrators, not the readers.

  Reading the documentation of cache I found references to fields of application different ehcache, but I don't know if these scopes also applies to the cachedwithin cached queries. Also, supposed to be the scope of default cache is server, moot...

  For example, queries are hidden with cachedwithin cached on a scope, as it seems? Is it possible to change it so better cached on a server scope? Would be better to migrate to cache such as cachePut() and cacheRemove() functions?

  Thanks in advance,

  (please forgive my bad English, some practices need more to write...)

  Cool, yes Rob Brooks-Bilson has added a comment to my blog article, pointing to the appropriate in the docs page.

  As well as doing the CFAdmin setting, we can also do in Application.cfc:

  Is not the default, but it redefines it to work the previous versions how the CF worked.

  --

  Adam

• Clear the cache with graphic flash

  Hi all
  
  On a menu, I have a graphic flash with a link to goto page xx
  The url in the link have the clear cache attributes defined for the same page xx.
  
  on the same page of the menu, I also provide a list with an option for even goto page xx
  f? p = 103:2:815028013919804:NO:2:
  
  
  Now the question
  When you use the menu list of the cache is disabled
  When you use the graphic flash cache is not authorized!
  
  Some body have an idea how it works?
  
  Hugo Perfors

  I think that it is a problem of IE. Try to change the caching there options to see if it happens after that. I think that if you are using FF it won't happen.

  Denes Kubicek
  -------------------------------------------------------------------
  http://deneskubicek.blogspot.com/
  http://www.Opal-consulting.de/training
  http://Apex.Oracle.com/pls/OTN/f?p=31517:1
  http://www.Amazon.de/Oracle-Apex-XE-Praxis/DP/3826655494
  -------------------------------------------------------------------

• Join between the tables with Order By

  Hello

  Sorry if my question has already been postponed. I Googled, but I can't find any solution.

  This is the version of my DB: database Oracle 10 g Enterprise Edition Release 10.2.0.3.0.

  I created 3 tables. They contain up to 4000 lines.

  When I join without the order of closed I got a very quick response (less than 1 second), when I simply add Order By it gets more than 30 seconds!

  The order is on a Date field from the first table.

  Here's my query:

  Select *.

  tableOne tone

  two tableTwo full outer join on tone.message_id = ttwo.message_id

  outer join complete tableThree on two.field_id = tthree.field_id

  tone.ts desc order;

  I have some difficulties to export the plan to explain it...

  Should I create any clue? where? on the orders of field? on the join field?

  Any tips? Thank you.

  NO, it would not be

  Oracle does not work like that.

  ORDER BY is always, really always done LAST.

  ALWAYS.

  The second statement is a JOIN EXTERNAL LEFT and not a FULL OUTER JOIN, so the results will be different.

  Why "assume you' yours 'assumptions' are always right?

  They are not. You were probably in your dipers when I was already performance in Oracle 6.0.33

  --------------

  Sybrand Bakker

  Senior Oracle DBA

• Get the list of the names of cache

  I use this code to get a list of all the names of my cluster cache:
  ------
  
  Cluster cluster = CacheFactory.ensureCluster ();
  for (enumeration < String > = cluster.getServiceNames services (); services.hasMoreElements () ;))
  {
  String sName = (String) services.nextElement ();
  System.out.println("service:"+sName);
  Service service = cluster.getService (sName);
  If (service instanceof CacheService) {}
  CService CacheService = service (CacheService);
  Caches of list < String > = cService.getCacheNames ();
  for (caches = cService.getCacheNames (); caches.hasMoreElements () ;){)
  Add caches.nextElement () in a list
  }
  }
  }
  
  ------
  
  It works fine, but in order to access multiple clusters, I have to use proxy/extend coherence.
  
  When I use the proxy/extend, the code above no longer works. I explain: when you use proxy/extend, cluster.getServiceNames () returns the Management and Cluster but does not return the Cache Services more.
  
  I tried to get a group in a different way object:
  
  . GetCluster (CacheFactory.getConfigurableCacheFactory () .ensureService ("ExtendTcpInvocationService") (InvocationService));
  
  where ExtendTcpInvocationService is defined with < remote-invocation-diet > but the result is the same.
  
  
  To summarize : I would like to get the list of all my names of cache in an application of several cluster; is this possible?

  Hello

  Yes, your invoked must be on the classpath of the server and you use POF, it must also be in the configuration of the server POF file.

  The best way to find the names of cache on the server without needing to code on the cluster must use JMX - assuming that you have enabled JMX on the cluster. You can easily run a JMX to Java code query or use something ike JConsole of VisualVM...

  JK

• Read-through Caching with time expired and nearly cache (foreground)

  We are experiencing a problem with our custom CacheLoader and near cache with expiration / delay on the storage card scheme.
  
  I was under the assumption that it was possible to have an expiry time configured on the system of support and that the cache near object was expelled during the backup object has been expelled. But, according to our tests, there must be a delay expires on the regime of front also.
  
  My hypothesis is correct, as there is not an automatic eviction on the cache near (foreground)?
  
  With this config, close cache is never cleared:

               <near-scheme>
                    <scheme-name>config-part</scheme-name>
                    <front-scheme>
                          <local-scheme />
                    </front-scheme>
                    <back-scheme>
                          <distributed-scheme>
                                <scheme-ref>config-part-distributed</scheme-ref>
                          </distributed-scheme>
                    </back-scheme>
              <autostart>true</autostart>
              </near-scheme>
  
  
              <distributed-scheme>
                    <scheme-name>config-part-distributed</scheme-name>
                    <service-name>partDistributedCacheService</service-name>
                    <thread-count>10</thread-count>
                    <backing-map-scheme>
                          <read-write-backing-map-scheme>
                                <read-only>true</read-only>
                                <scheme-name>partStatusScheme</scheme-name>
                                <internal-cache-scheme>
                                      <local-scheme>
                                            <scheme-name>part-eviction</scheme-name>
                                            <expiry-delay>30s</expiry-delay>
                                      </local-scheme>
                                </internal-cache-scheme>
                                <cachestore-scheme>
                                      <class-scheme>
                                            <class-name>net.jakeri.test.PingCacheLoader</class-name>
                                      </class-scheme>
                                </cachestore-scheme>
                                <refresh-ahead-factor>0.5</refresh-ahead-factor>
                          </read-write-backing-map-scheme>
                    </backing-map-scheme>
                    <autostart>true</autostart>
                    <local-storage system-property="tangosol.coherence.config.distributed.localstorage">true</local-storage>
              </distributed-scheme>

  With this config (added expiration-behind front-scheme), near cache gets cleared.

          <near-scheme>
                    <scheme-name>config-part</scheme-name>
                    <front-scheme>
                          <local-scheme>
                               <expiry-delay>15s</expiry-delay>
                          </local-scheme>
                    </front-scheme>
                    <back-scheme>
                          <distributed-scheme>
                                <scheme-ref>config-part-distributed</scheme-ref>
                          </distributed-scheme>
                    </back-scheme>
              <autostart>true</autostart>
              </near-scheme>
  
  
              <distributed-scheme>
                    <scheme-name>config-part-distributed</scheme-name>
                    <service-name>partDistributedCacheService</service-name>
                    <thread-count>10</thread-count>
                    <backing-map-scheme>
                          <read-write-backing-map-scheme>
                                <read-only>true</read-only>
                                <scheme-name>partStatusScheme</scheme-name>
                                <internal-cache-scheme>
                                      <local-scheme>
                                            <scheme-name>part-eviction</scheme-name>
                                            <expiry-delay>30s</expiry-delay>
                                      </local-scheme>
                                </internal-cache-scheme>
                                <cachestore-scheme>
                                      <class-scheme>
                                            <class-name>net.jakeri.test.PingCacheLoader</class-name>
                                      </class-scheme>
                                </cachestore-scheme>
                                <refresh-ahead-factor>0.5</refresh-ahead-factor>
                          </read-write-backing-map-scheme>
                    </backing-map-scheme>
                    <autostart>true</autostart>
                    <local-storage system-property="tangosol.coherence.config.distributed.localstorage">true</local-storage>
              </distributed-scheme>

  Hi Jakkke,

  Plan near Cache allows configurable levels of most basic cache coherence cache expiration based versioning of data cache according to the requirements of consistency of the cache invalidation based. Near Cache is commonly used to achieve the performance of the cache replicated without losing aspects of scalability of the replicated cache, and this with a subset of data (based on SRM or UPM) in the of the near cache and all the data in the cache close. Updates of can automatically trigger events to invalidate the entries in the invalidation strategy-based (present, all, none auto) configured for the cache close.

  If you want to expire the entries in the and , you must specify a time-out on the two schemes as mentioned by you in the last example. Now if you are to expire items in the for the reason that they get loaded from the cache-store but remain even keys (only the values should be updated in the cache store), so you should not set the expiration delay on the mention rather the invalidation strategy as a gift. But if you want to have a different set of entries in after a specified delay so that you need to mention this in the configuration of .

  Near cache has the ability to keep the plan before and backup data system synchronized, but the expiration of the entries is not synchronized. Still, avant-regime is a subset of background.

  I hope this helps!

  See you soon,.
  NJ

• Unable to create cache with several tables group

  Hello
  
  I need to create a group of cache with several tables, which are referential to the other.
  
  There are 2 related tables and table 1 child...
  While trying this thing, it gives me an error:
  
  8222: multiple parent tables found
  
  It is not possible to use several tables of root in a single cache group? Is there another way to use it?
  
  Script, I used is:
  
  create a cache group asynchronous writethrough TEST. CG1
  of the TEST. ROOT1)
  KEY PRIMARY ID VARCHAR2 (8 BYTES),
  NAME VARCHAR2 (50 BYTE),
  DESCRIPTION VARCHAR2 (255 BYTE),
  POLICYTYPEID VARCHAR2 (7-BYTE)),
  
  TEST. ROOT2)
  PARAMTYPEID VARCHAR2 (5 BYTES) PRIMARY KEY,
  PARAMETERUSAGE VARCHAR2 (1 BYTE),
  NAME VARCHAR2 (25 BYTE),
  DESCRIPTION VARCHAR2 (255 BYTE)),
  
  TEST. CHILD1)
  PARAMDETAILID NUMBER (20) PRIMARY KEY,.
  ID VARCHAR2 (8 BYTE),
  PARAMETERUSAGE VARCHAR2 (1 BYTE),
  DISPLAYVALUE VARCHAR2 (255 BYTE),
  OPERATORID VARCHAR2 (5 BYTE),
  VENDORID NUMBER (20).
  FOREIGN KEY REFERENCES TEST. ROOT1 (ID),
  TEST KEY (PARAMETERUSAGE) REFERENCES STRANGERS. ROOT2 (PARAMETERUSAGE));

  You can't have multiple root within a group of cache tables. The requirements for tables in the group a cache are very strict; There must be only one top-level (root table) table and there may possibly be several children tables. Tables of the child should be linked through foreign keys to the root table or a child table above in the hierarchy.

  The solution for your case is among the tables of root and the other root table in a separate cache group and the child table in a cache group. If you do this, you must take care of a few things:

  1. you cannot define foreign keys between the tables of groups of different cache in TimesTen (keys may exist in Oracle) so the application must enforce referential integrity itself for these cases.

  2. If you load data in a cache group (using LOAD the GROUP CACHE or "load we demand") and Timesten will not automatically load the corresponding data in the other group of cache (since he doesn't know the relationship). The application must load the data into the other group of cache explicitly.

  There is no problem with transactional consistency when changes are pushed to Oracle. TimesTen maintains and reinforces the coherence transactional regardless of how the tables are arranged in groups cache correctly.

  Chris