Database network packet size

How can one change the database network packet size in Oracle 8i in the tnsnames file?

Search "SDU" in the documentation. Fake delirium about size "tuning" - set SDU to the maximum 32767. TCP stack much mor clever than Oracle's developers can explain.

Similar Messages

Network Packet Size Or Bandwidth Restriction in Database Mirroring

Hi Everyone,
I tried finding online but couldn't get answer to the below question.
Is it possible to restrict the packet size for the SQL Database Mirroring? Is it possible to tell database mirroring what size packets it should send to the mirrored database? Is it possible to define the bandwidth or packets per second for mirroring? If
yes, is it possible at the server level or database level? If yes, how to do this?
The reason that I want to know this is because, if I have N number of databases on a server and I'm mirroring them on other server in a separate location, how will I get to know the bandwidth constraint and if there is a bandwidth constraint can I control
the data sent through database mirroring to ensure that the bandwidth is not affected?
Thanks
Mayur Kashikar

Database Mirroring has no such setting specific to "per database" or "per server" which can limit the packets sent over the wire. Remember that communication of database doesn't happen via regular port. There is a separate port use by
database mirroring (generally 5022)
Network packet size is setting which is for communication happening on regular connection port (i.e. TDS packets). Database mirroring doesn't use TDS packets but it uses the Service Broker network stack.
In short, there is no such setting which can control the communication the way you want.
If you search for Jumbo Frame, you might get some reference about performance gain but remember that if you use Jumbo frame then it has to be supported by each and every component in the network else you would have packet drop and mirroring would be
disconnected.
Balmukund Lakhani
Please mark solved if I've answered your question, vote for it as helpful to help other users find a solution quicker
This posting is provided "AS IS" with no warranties, and confers no rights.
My Blog |
Team Blog | @Twitter
| Facebook
Author: SQL Server 2012 AlwaysOn -
Paperback, Kindle

MDW collection failing due Network Packet Size

Need a work around for MDW collection. Network pkt size gt then 16,383 bytes.http://tinyurl.com/mc9nzkq
I know reducing network packer size will resolve the issue . But I am looking for a work around if anybody knows.
thanks,

Hi Taiob,
Can you provide more information on:
Which system collectors you're using, if any
The query for a custom collector, if you're using one and you can provide it
the edition of SQL Server you're using
Thanks,
George Li
Program Manager, SQL Server

Setting maximum packet size in JDBC driver to send data to database

Could someone tell me how I can set the JDBC connection property of maximum packet size to send data to database?
Regards
Rashed

Hi thanks....I'm having this strange SQLException while trying to insert BLOB image data to Oracle database. I'm saying this strange because for the same image that has been inserted before it's throwing the exception. My program is run from Oracle form and then some image data are inserted into database through a loop. I can't realize what's the problem inside my code that's causing this problem. In fact, when I run my program independently not from Oracle Form, it runs fine, every image data get inserted into database. Given below is my code snippet:
public void insertAccDocs(String[] accessions) throws SQLException
 for(int q=0; q<accessions.length; q++)
 final String docName = accessions[q];
 dbThread = new Thread(new Runnable(){
 public void run()
 try{
 System.out.println("insertDB before connection");
 getConnected();
 System.out.println("insertDB after connection");
 st=con.createStatement();
 //String docName = acc; commented
 // String docName = singleAccession;
 String text = formatFree;
 String qry = "INSERT INTO DOCUMENT VALUES
('"+docName+"','"+text+"','"+formatted+"','"+uiid+"')";
 System.out.println("parentqry"+qry);
 int ok=0;
 ok=st.executeUpdate(qry);
 if(ok==1)
 System.out.println("INSERTION SUCCESS= "+ok);
 System.out.println("Image List Size"+ imgList.size());
 // inserting into child
 for(int i=0;i<imgList.size(); i++)
 String imgPath = ""+imgList.get(i);
 System.out.println("db"+imgPath);
 FileInputStream fin = new FileInputStream(imgPath);
 BufferedInputStream bufStr = new BufferedInputStream(fin);
 byte[] imgByte = new byte[bufStr.available()];
 String img = "" + imgNameList.get(i);
 System.out.println("imgid="+i);
 callable = con.prepareCall("{call prc_insert_docimage(?,?,?)}");
 callable.setString(1,docName);
 callable.setString(2,img);
 callable.setBinaryStream(3, bufStr , (int)imgByte.length);
 callable.execute();
 callable.close();
 con.commit();
 con.close();
 else
 System.out.println("INSERTION NOT SUCCESS");
 // } //else end of severalAcc
 }catch(Exception err){ //try
 System.out.println(err.toString());
 } //run
}); //runnable
dbThread.start();
}And the exception thrown is :
java.sql.SQLException: Data size bigger than max size for this type: #####
Please let me know if possible how I can get around this.
Regards
Rashed

How u will set the external parameters like packet size and number

How u will set the external parameters like packet size and number of parallel process.

Dear Karthik,
ALE:
Use Transaction UPSC02 - ALE Distribution Unit: Packet Types.
Performance Optimization for ODS Objects:
To ensure a good ODS object loading performance, take the following into account:
 1. Creating SIDs
The creation of SIDs takes a long time and can be avoided in the following cases by:
 Not setting the indicator for BEx Reporting if you are using the ODS object only as a data store. If you do set this indicator, SIDs are created for all new characteristic values.
 If you use line items (for example, document number or time stamp) as characteristics in the ODS object, in the characteristic maintenance, indicate these as Attribute Only.
SIDs are created in parallel if parallel activation is switched on (see last point). They are then created with the same number of parallel processes as created for the activation.
However, if you specify a server group or a special server in customizing, these specifications are not only valid for the activation, but also for the SID creation. The creation of SIDs runs on the application server on which the batch job also runs.
 2. DB partitioning in active data tables (technical A table)
By partitioning by database level, you can delete data from the ODS object much more quickly. As a partitioning criterion, choose the characteristic by which you want to delete. For more information on partitioning database tables, see the database documentation (DBMS-CD). Partitioning is supported by the following databases: Oracle, DB2/390, Informix.
 3. Indexing
Use selection criteria for queries for ODS objects. If the key fields are specified, the existing primary index is used. The more frequently accessed characteristic should appear on the left-hand side.
If you did not specify the key fields completely in the selection criteria (visible in the SQL trace), you improve the run time of the query by creating additional indexes. You can create these secondary indexes in the ODS object maintenance.
 4. Activation of data in an ODS object
To improve system performance when activating data in the ODS object, you can make the following entries in Customizing under SAP Customizing Implementation Guide -> SAP NetWeaver -> Business Information Warehouse -> General BW Settings -> ODS Object Settings:
 the maximum number of parallel processes when activating data in the ODS object as when moving SIDs
 the minimum number of data records for each data package when activating data in the ODS object, meaning you define the size of the data packages that are activated
 the maximum wait time in seconds when activating data in the ODS object. This is the time when the main process (batch process) waits for the dialog process that is split before it classifies it as having failed.
 the server group that needs to be used when activating the data in ODS objects in parallel You have to create the server groups beforehand using the following path: Tools -> Administration -> Network -> RFC Destination, RFC -> RFC Groups. If you do not specify anything here, then the activation runs on the server on which the batch process was started for activation. If a server from the server group is not active, then the activation is cancelled.
 5. Loading unique data records
If you only load unique data records (meaning data records with nonrecurring key combinations) into the ODS object, then the loading performance is improved when you set the indicator Unique Data Records in the ODS object maintenance.
The records are then updated more quickly because the system no longer needs to check whether the record already exists. You do have to be sure that there duplicate records are definitely not being loaded, because this will lead to termination.
Regards,
Naveen.

DNS after reducing the advertised EDNS UDP packet size to 512 octets

hello
There is something wrong with my DNS server, it open internet webs so slow,and i have no idea with this .
04-Mar-2011 14:31:57.264 zone 0.0.127.in-addr.arpa/IN/com.apple.ServerAdmin.DNS.public: loaded serial 1997022700
04-Mar-2011 14:31:57.264 zone 15.0.168.192.in-addr.arpa/IN/com.apple.ServerAdmin.DNS.public: loaded serial 2011030204
04-Mar-2011 14:31:57.265 zone ******/IN/com.apple.ServerAdmin.DNS.public: loaded serial 2011030404
04-Mar-2011 14:31:57.265 zone localhost/IN/com.apple.ServerAdmin.DNS.public: loaded serial 42
04-Mar-2011 14:31:57.265 running
04-Mar-2011 14:32:00.066 host unreachable resolving 'b.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:7fd::1#53
04-Mar-2011 14:32:00.261 host unreachable resolving 'r.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:500:2f::f#53
04-Mar-2011 14:32:00.261 host unreachable resolving 'r.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:503:c27::2:30#53
04-Mar-2011 14:32:00.261 host unreachable resolving 'r.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:503:ba3e::2:30#53
04-Mar-2011 14:32:00.261 host unreachable resolving 'r.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:7fd::1#53
04-Mar-2011 14:32:00.261 host unreachable resolving 'r.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:dc3::35#53
04-Mar-2011 14:32:00.361 host unreachable resolving 'dr.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:500:1::803f:235#53
04-Mar-2011 14:32:00.361 host unreachable resolving 'dr.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:7fd::1#53
04-Mar-2011 14:32:00.361 host unreachable resolving 'dr.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:dc3::35#53
04-Mar-2011 14:32:00.361 host unreachable resolving 'dr.dns-sd.udp.0.0.168.192.in-addr.arpa/PTR/IN': 2001:503:ba3e::2:30#53
04-Mar-2011 14:32:00.866 host unreachable resolving './NS/IN': 2001:7fd::1#53
04-Mar-2011 14:32:01.005 success resolving './NS' (in '.'?) after reducing the advertised EDNS UDP packet size to 512 octets
04-Mar-2011 14:32:01.408 success resolving 'local/SOA' (in '.'?) after disabling EDNS
04-Mar-2011 14:32:01.533 host unreachable resolving 't.arin.net/AAAA/IN': 2001:503:ba3e::2:30#53
04-Mar-2011 14:32:01.534 host unreachable resolving 'v.arin.net/AAAA/IN': 2001:500:1::803f:235#53
04-Mar-2011 14:32:01.534 host unreachable resolving 'v.arin.net/AAAA/IN': 2001:7fd::1#53
04-Mar-2011 14:32:01.534 host unreachable resolving 'v.arin.net/AAAA/IN': 2001:503:ba3e::2:30#53
04-Mar-2011 14:32:01.534 host unreachable resolving 'w.arin.net/A/IN': 2001:500:1::803f:235#53
anyone help!! Thanks a lot.
Message was edited by: leo.xue

also interesting, when setting DNS servers IPs in Network Settings for your host,
it is important that the IPs are not doubled.
in example if you have Networksettings like
automatic DHCP given from router or server.
IP-Adress: 192.168.2.2
Subnetmask: 255.255.255.0
Router IP: 192.168.2.1
DNS-Server: 127.0.0.1, 192.168.2.2
wich means localhost and again same machine, just different IP..
then your lookup mechanism has to walk thru this steps to know if there is nothing inside to resolve adresses.
which means in this example it would take double the time if DNS-Server would be just 127.0.0.1
you can see if there is a lot to work thru in your logs.
look for something like "sizing zone task pool based on 9 zones".
this mount of zones will change if you set the correct DNS server IP.
more zones are slower than less, very logical!
but this will not solve your problem with packet size at all, it just reduces circles after dns resolves not known adresses even with packet size change.

T3/RMI packet size very large

In order to determine our bandwidth requirements, we recently placed a
sniffer on our network to analyze our message packets. We've noticed
that
our packets are very large and it appears that much of the overhead is
in
the overhead of RMI or T3.
Here are some sample numbers for similar message between a single client
and our WLS server.
T3: 3500 bytes
T3 w/ HTTP tunneling: 5500 bytes
IIOP: 1250 bytes (using VisiBroker ORB talking to Smalltalk ORB)
As you can see the T3 packet size is 65% larger than the same packet
sent
via Corba/IIOP. It also appears that with RMI, all of the full class
names
and variable names are also being passed along within the packet. Are we
missing something or is this an understood fact? Is there anything we
can
do to fix this problem? As it stands, our bandwidth requirements to
support
the larger T3 packet size are astronomical and this would not be
feasible
in a production environment. Does anyone know what is the typical
percentage
overhead increase per packet. It appears to be about 400%.
Our WLS environment is described below.
Edwin Marcial
Continental Power Exchange
Weblogic Environment
WLS Server
WLS 4.51 w/ Service Pack 7
NativeIO = true
ExecuteThreadCount = 40
readTimeoutMillis=5000
readTimeoutMillisSSL=10000
Dell Pentium III 600 w/ 512 MB memory
JavaSoft 1.2.2
-ms128 -mx350
WLS Client
Java Application
t3s and https (using WLS RMI)
JavaSoft 1.1.7b
typically Pentium 200 MHz or better w/ 64MB or more

I think you are kind of stuck with this. RMI is a heavyweight protocol in
comparision to IIOP. If the message sizes really bother you that much you
may want to look into and a EJB implementation that maps RMI to IIOP such as
the Inprise Application Server which sits atop the VisiBroker ORB.
-paul
Edwin Marcial <[email protected]> wrote in message
news:[email protected]..
In order to determine our bandwidth requirements, we recently placed a
sniffer on our network to analyze our message packets. We've noticed
that
our packets are very large and it appears that much of the overhead is
in
the overhead of RMI or T3.
Here are some sample numbers for similar message between a single client
and our WLS server.
T3: 3500 bytes
T3 w/ HTTP tunneling: 5500 bytes
IIOP: 1250 bytes (using VisiBroker ORB talking to Smalltalk ORB)
As you can see the T3 packet size is 65% larger than the same packet
sent
via Corba/IIOP. It also appears that with RMI, all of the full class
names
and variable names are also being passed along within the packet. Are we
missing something or is this an understood fact? Is there anything we
can
do to fix this problem? As it stands, our bandwidth requirements to
support
the larger T3 packet size are astronomical and this would not be
feasible
in a production environment. Does anyone know what is the typical
percentage
overhead increase per packet. It appears to be about 400%.
Our WLS environment is described below.
Edwin Marcial
Continental Power Exchange
Weblogic Environment
WLS Server
WLS 4.51 w/ Service Pack 7
NativeIO = true
ExecuteThreadCount = 40
readTimeoutMillis=5000
readTimeoutMillisSSL=10000
Dell Pentium III 600 w/ 512 MB memory
JavaSoft 1.2.2
-ms128 -mx350
WLS Client
Java Application
t3s and https (using WLS RMI)
JavaSoft 1.1.7b
typically Pentium 200 MHz or better w/ 64MB or more

Is there any limitaion for JDBC packet size?

Hi there,
my question is if JDBC like ODBC have the 4k
packet size? If so, is there a way to increase it and if so, what will be the maximum acceptable size for it?
Thanks a lot.
null

Oracle supports JDBC at least as far back as 8.1.something; the Oracle JDBC driver is built on a private protocol that existed in at least 8.0, so I'd expect all 8x versions to support JDBC, more or less.
http://www.oracle.com/technology/tech/java/sqlj_jdbc/htdocs/jdbc_faq.html
http://www.oracle.com/technology/sample_code/tech/java/sqlj_jdbc/index.html
http://www.oracle.com/technology/software/tech/java/sqlj_jdbc/index.html
However, you may want to consider using a newer version of the Oracle database; you can download the latest, and a single-user-for-development-only license is free.
http://www.oracle.com/technology/software/products/database/oracle10g/index.html

Controlling packet size from 32K to 4K for SQL DSN

I've been engaged by my SQL administrator to find out how the packet size can be reduced from 32K to 4K in the CF9 Enterprise environment in a multi-server configuration. Although it's not likely related to the version rather than the JDBC Drivers provided in the product, none the less there is no clear way of identifying where this could be changed within the ColdFusion DSN's. Does anybody have any insight on how to control the packet sizes in a SQL DSN i Coldfusion 9 or 9.0.1? Would be greatly appreciated.

If you are using MSSQL, open the CF administrator, goto the datasource, click the "Advanced Settings" button and enter in:
packetSize=8
within the Connection String textarea box then hit submit, there should be no need to restart anything
Connection details for the PacketSize property:
Valid Values
-1 | 0 | x
where x is an integer from 1 to 128 that represents a number of bytes.
If set to -1, the driver uses the maximum packet size that the database server accepts.
If set to 0, the driver uses the default packet size configured on the database server.
If set to x, the driver uses a packet size that is calculated using the specified value multiplied by 512.
Example
If PacketSize=8, the packet size is set to 8 * 512 bytes (4096 bytes).
Default
-1
Data Type
int
Microsoft recommends (4096 bytes)

Regrd the data packet size

Hi
i have two questions
1) where can i set the data packet size in BW and R3 and also in BI how and where can we set .
2)by what method/logic we will select the key fields in dso.
ex: i have 5 tables in the sourse and each table will have primary keys, now how do we know that particular primary keys should be kept in KEY FIELDS folder in DSO.
full points will be assigned.

HI,
Data package settings for the data to be extracted from R3 to BI can be done through :
1) SBIW>General Settings> Maintain Control Parameters for Data Transfer
These settings are common for all the info packages which extract data from R3.
2) If u want to do settings relevent to specific Infopackage then :
RSA1>Click On the Specific Infopackage>Scheduler(in the Menu Bar)-->DataS Default Data Transfer.
3) And if you want to do DSO Package settings then:
Got to Transaction RSODSO_SETTINGS
Here u can do package settings for DSO activation ,paramenter for SID generation ect.
And selection of Key fileds depends upon the requirement.
Based on the key fields what all the data fileds u want to overwrite or add the corespomding data fields
Regards,
Chaitanya.

How do I change packet size for a single load

All,
I need to change the packet size for a single load I'm doing. How do I set that?
Thanks.
Dave

Hi..........
For this u hav to go to the infopackage scheduler........
If u r in Process chain >> Right click on the infopackage >> Maintain variant..........then it will come to the IP scheduler..................In the Top Scheduler Tab >> Click on it and select DataS. Default data transfer .......Now there may be a delta IP and Full upload IP for the same datasource.......check ur IP .....it is full or delta..........if full change the packet size for full upload..............otherwise for delta upload........
If the IP is not a part of any Process chain..........then go to RSA1 >> Find the IP..........and double click on it........then the IP scheduler screen will open..........
You can make global settings for all datasources...... in the transaction SBIW in the area General Settings >> Maintain control Parameter for Data transfer......... The parameters that you maintain here are valid as global standard values, and thus apply to all DataSources.
Hope this helps.....
Regards,
Debjani........
Edited by: Debjani Mukherjee on Nov 13, 2008 2:19 PM

To list only user databases with the size for a instance in sql server 2005

Hi,
I looking for T-SQl to list only user databases with their size for a instance in sql server 2005

Try this:
use [databasename]
go
if convert(varchar(20),SERVERPROPERTY('productversion')) like '8%'
SELECT [name], fileid, filename, [size]/128.0 AS 'Total Size in MB',
[size]/128.0 - CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0 AS 'Available Space In MB',
CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0 AS 'Used Space In MB',
(100-((([size]/128.0 - CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0)/([size]/128.0))*100)) AS 'percentage Used'
,((CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0)/([size]/128.0))*100 as 'percentage used'
FROM sysfiles
else
SELECT [name], file_id, physical_name, [size]/128.0 AS 'Total Size in MB',
[size]/128.0 - CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0 AS 'Available Space In MB',
CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0 AS 'Used Space In MB',
(100-((([size]/128.0 - CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0)/([size]/128.0))*100)) AS 'percentage Used'
--,((CAST(FILEPROPERTY(name, 'SpaceUsed') AS int)/128.0)/([size]/128.0))*100 as 'percentage used'
FROM sys.database_files
go
Or can refer below link:
http://gallery.technet.microsoft.com/scriptcenter/All-Databases-Data-log-a36da95d
Please click the Mark as answer button and vote as helpful if this reply solves your problem

Packet size and selection screen

hi all.,
i have a requirement where user just need to give the no. of records to be processesed for a execution of a report.
i am using packet size were the user gives the no. of records he want to be executed in a report.
Please let me know how to use packet size in my selection screen
thnkx
bhanu

hi,
check this code regarding packagr size ,no of records to be processed to get out put in report,
TABLES : mkpf,
 mseg,
 zamatdoc,
 mara.
DATA :BEGIN OF gi_mkpf OCCURS 100,
 mblnr LIKE mkpf-mblnr,
 mjahr LIKE mkpf-mjahr,
 END OF gi_mkpf.
data gi_mseg like zamatdoc occurs 100 with header line.
DATA: BEGIN OF gi_mseg OCCURS 0.
 INCLUDE STRUCTURE zamatdoc.
DATA: END OF gi_mseg.
DATA: gv_date(11)," LIKE ekpo-aedat,
 gv_date1(11),
 year(4),
 month(2),
 day(2),
 flag(1),
 read_flag(1).
DATA :char TYPE i,
 count TYPE i.
*select-options : s_mtart for mara-mtart obligatory.
SELECTION-SCREEN: BEGIN OF BLOCK b1 WITH FRAME TITLE text-110.
SELECT-OPTIONS : records FOR char. " OBLIGATORY.
SELECTION-SCREEN : END OF BLOCK b1.
START-OF-SELECTION.
gv_date1 = sy-datum.
year = gv_date1+0(4).
month = gv_date1+4(2).
day = gv_date1+6(2).
year = year - 3.
CLEAR gv_date1.
CONCATENATE year month day INTO gv_date1.
CONCATENATE day month year INTO gv_date1.
SELECT mblnr
 mjahr
 FROM mkpf
 INTO TABLE gi_mkpf
 WHERE budat <= gv_date1.">" and BUDAT >= ).
IF sy-subrc = 0.
 SORT gi_mkpf BY mblnr.
 DELETE ADJACENT DUPLICATES FROM gi_mkpf.
 CLEAR count.
 IF NOT gi_mkpf[] IS INITIAL.
 IF NOT records-low IS INITIAL .
 IF NOT records-high IS INITIAL.
 DESCRIBE TABLE gi_mkpf LINES count.
 IF records-high < count.
 DELETE gi_mkpf FROM records-high TO count.
 ENDIF.
 IF records-low <> 1.
 IF records-low <> 0.
 DELETE gi_mkpf FROM 1 TO records-low.
 ENDIF.
 ENDIF.
 ENDIF.
 ENDIF.
 ENDIF.
SELECT matnr werks FROM mseg INTO CORRESPONDING FIELDS OF TABLE gi_mseg
 PACKAGE SIZE 100 FOR ALL ENTRIES IN gi_mkpf
 WHERE mblnr = gi_mkpf-mblnr
 AND mjahr = gi_mkpf-mjahr.
 MODIFY zamatdoc FROM TABLE gi_mseg.
 COMMIT WORK.
 ENDSELECT.
regards
siva

CWMS 2.0 Error:unsupported Configuration-IRP not Receiving Network Packets

Hi,
I am having a 50 user deployment non redundant setup using non split DNS.The version used for deployment is CWMS 2.0.1.107B-AE .I have configured the ports as per the planning guide for manual deplyment,public access and Non split DNS. Forward and Reverse DNS Entries for are also configured
After deploying the vmmachine at the end ,during the system check i get an error showing that the irp in the DMZ is not able to receive network packets and thus the deployment status is shown as Unsupported Configuration
Attaching the error received. How do I troubleshoot this?Has anyone come across this error?

Hi,
You are definitely experiencing a network connectivity issue between the Admin and IRP VMs. Please, ensure that PING (ICMP) is allowed between the Admin and IRP VMs as well as all other required ports (http://www.cisco.com/c/en/us/td/docs/collaboration/CWMS/2_0/Planning_Guide/Planning_Guide_chapter_011.html#reference_CB65D7FE4B3746DDAF1649884AD777CE). Also, ensure that DNS server is reachable from internal and DMZ networks.
Furthermore, I can see that you are using domain.internal for your VM domains. Per Certification Authority announcements, they will stop issuing SSL certificates for internal domains next year (http://www.digicert.com/internal-names.htm) so it might be good to assign a valid domain name for your VMs. This is not related to the current issue you are seeing, but can help you avoid issues with SSL certs in the future.
-Dejan

Default maximum packet size in 3.7.0.1

Hello,
It looks like there is another reason for performing a full cluster restart to apply the first patch for 3.7 (3.7.0.1). On a machine with MTU=1500, it appears that the default maximum packet size value was changed from 1452 to 65535 and this change prevents the cluster from forming during a rolling upgrade.
16 May 2011 14:08:01,376 ERROR Logger@813856187 3.7.0.1: 2011-05-16 14:08:01.376/1.539 Oracle Coherence GE 3.7.0.1 <Error> (thread=Cluster, member=n/a): This member could not join the cluster because of a configuration mismatch between this member and the configuration being used by the rest of the cluster. The maximum packet size (65535) for this member does not match the maximum packet size that the running cluster is using. Rejected by Member(Id=1, Timestamp=2011-05-16 14:05:49.303, Address=XXX2:11500, MachineId=38601, Location=machine:YYY2,process:14945,member:cacheserver:1, Role=cacheserver).
16 May 2011 14:24:27,675 ERROR Logger@98371294 3.7.0.0: 2011-05-16 14:24:27.674/1.539 Oracle Coherence GE 3.7.0.0 <Error> (thread=Cluster, member=n/a): This member could not join the cluster because of a configuration mismatch between this member and the configuration being used by the rest of the cluster. The maximum packet size (1452) for this member does not match the maximum packet size that the running cluster is using. Rejected by Member(Id=1, Timestamp=2011-05-16 14:11:11.927, Address=XXX1:11500, MachineId=38600, Location=machine:YYY1,process:18509,member:cacheserver:1, Role=cacheserver).
Can you please confirm?
Cheers,
Alexey

Hi Alexey,
Yes, the default maximum packet size for 3.7 was supposed to have been 65535, there was however a bug which prevented this from working. 3.7.0.1 fixes this issue, and thus "introduces" the new higher default value. A rolling restart is still possible, you would however need to manually configure the maximum to be 1468. The new maximum of 65535 allows the Coherence default configuration to support large clusters (greater then ~450 JVMs), where as the old default configuration would prevent new members from joining and would require a re-configuration and full restart.
thanks,
Mark
Oracle Coherence

Database network packet size

Similar Messages

Maybe you are looking for