Unicast clustering option
Is it possible to implement the clustering using unicast?
I would think if implemented on top of the tcp layer, you could get rid of the
reliability code to handle the udp of multicast.
Would it be that hard to implement the proxy with normal tcp/ip?
Is it necessary to make the clustering feature dependent on a network capability
that a lot of IT shops won't enable, or the piece of infrastructure it requires
is in a different department than the Java people?
I was speaking to a network guy, he said he was not aware of the use of multicast
in a LAN for the purpose of clustering, he thought it was a little strange dependency
given the amount of bandwidth you get in the LAN. He noted that multicast in
general is for a different problem (audio/video over the web).
Best regards,
-R
Is it possible to implement the clustering using unicast? >
> I would think if implemented on top of the tcp layer, you could get rid of
the
> reliability code to handle the udp of multicast.
>
> Would it be that hard to implement the proxy with normal tcp/ip?
>
> Is it necessary to make the clustering feature dependent on a network
capability
> that a lot of IT shops won't enable, or the piece of infrastructure it
requires
> is in a different department than the Java people?
>
> I was speaking to a network guy, he said he was not aware of the use of
multicast
> in a LAN for the purpose of clustering, he thought it was a little
strange dependency
> given the amount of bandwidth you get in the LAN. He noted that multicast
in
> general is for a different problem (audio/video over the web).
There is a difference between Internet multicast (video over the Internet
for example) and the multicast that Weblogic uses. Multicast as a technology
on TCP/IP networks predates the concept of video on the Internet by a long
shot, and has been utilized for cluster-like processing for many years. See
PGM (http://www.networksorcery.com/enp/rfc/rfc3208.txt) as one example.
If you don't like how WebLogic does multicast, you can do your own multicast
routing by writing a program to listen on that multicast address and use
TCP/IP to known IP/ports for point-to-point communication. Then set the TTL
to 0 and you've removed multicast from the picture. (We have a prototype
that is similar for doing clustering between geographically distributed
sites over the public Internet.)
Peace,
Cameron Purdy
Tangosol, Inc.
http://www.tangosol.com/coherence.jsp
Tangosol Coherence: Clustered Replicated Cache for Weblogic
"r. kucera" <[email protected]> wrote in message
news:[email protected]...
>
Similar Messages
-
Unicast Clustering in Weblogic 10.3
We are using Multicast with Weblogic 9.2 for clustering now but are now migrating to Weblogic 10.3 and want to switch to unicast. The documentation seems kind of hazy for Unicast clustering. I have a few questions.
1)The Unicast Broadcast Channel. What exactly is it and what is the format of one.
2) do I need to create a custom channel some where? I don't want to use the default channel because we have many domains in the same server.
3) Generally has does Unicast work and how is it different from multicast.
Thanks,
Tom.Hi Tom
This really helped me
Ash -
Mail clustering options greyed
On a clean/default Lion Server install, I see this, when I press the Change... buton next to the Clustering option in Server Admin:
Anyone know why the bottom two options would be greyed out?
RustyHi
AFAIK in previous versions of OS X Server Mail Clustering was not supported unless you had a SAN product. I could be wrong but I don't think it would work with anything other than XSAN?
http://manuals.info.apple.com/en_US/MailServicesAdmin_v10.6.pdf
Page 82.
For lack of any thing else coming from Apple themselves it looks like this is still the case with Lion Server? I think the best advice I could give you is to either wait and see if forthcoming updates changes available functionality, or revert back to Snow Leopard Server and wait until Lion Server becomes operational in the way it is hoped it will be. As well as waiting for any further documentation to be made available.
You may also want to consider other Mail Server products instead?
HTH?
Tony -
Clustering options for FirePOWER SSP-40
What are the option for clustering Firepower SSP-40?
ASA 5585X with 2 node cluster, is it possible to have a single cluster for IPS service also ?
Concers are , how is the IPS module handling the Asymmetric situation...?
Cisco ASA 5585-X with FirePOWER SSP-40..What Cisco was missing imo was the NG Firewall features that everyone has had for years. I recently upgraded my 5585 SSP60 based firewalls with the SSP-SFR60 modules. So far I'm fairly impressed, the package does what everyone else does, currently still testing a number of items however what I could do with a Palo or something from SonicWall I can now do with my ASAs.
First Accessible from Same GUI: Not on the 5585s, if I open ASDM for the 5585s, there are no configurations that I can see available, essentially under my admin context I see three tabs, one of which is ASA Firepower Status, there is a link I can click on and it takes me to DC URL. I do hear you can configure for the smaller firewalls though. Possibly because I use MC and A/A I might be missing it. Honestly though DC seems to be a very powerful tool so I'm fine with two management systems as they both serve a different purpose, ASA controls ports allowed to pass, and the traffic that passes gets filtered.
Logging and Events: DC has an immense amount of information, urls being visited, ips, responding countries, pretty much everything is here.
Am I satisfied? Yes, items that bothered me was built in URL filtering, file analysis, Geo Location Filtering and numerous other bells and whistles that I noticed $1000 sonicwalls had, the CX platform never caught my eye for some reason, and I'm glad I never bit the bullet, EOL10154, so if I implemented CX I'd be in a situation where I would need to replace it. Sourcefire was a fairly successful company that Cisco bought, and I don't see cisco throwing away 2.7 Billion Dollars so I can see this being around for awhile.
I'm still exploring but so far what everyone else has had for years, I finally have with my ASA and I'm happy. It does work, and it works well. -
When to use the -clusterable extra rmic option?
Hi,
We are trying to do a clustered deployment on the WLS 701. I am trying to figure
out all the things that need to be on for clustering to work successfully. Currently
I am focused on the EJBs. Here is what I have found:
o For remote SLSBs there is no need to specify home-is-clusterable in weblogic-ejb-jar.xml
because the default value is true?
o For remote SLSBs there is no need to specify is-clusterable in weblogic-ejb-jar.xml
because the default value is true?
o For Entity Beans is-clusterable doesn't make sense?
o For local Entity Benas home-is-clusterable doesn't make sense?
o Are the options for home-is-clusterable and is-cluterable enough or do we have
to specify -clusterable as extra rmic option under ejbc options?
Thank you so much,
-Bart Simpson
Hi,
How are you doing? Is anybody out there using the -clusterable
option? Why?
Does the option make sense even if the both the JSP/Servlets and EJBs
are deployment on the name node?
Thank you so much,
-Bart Simpson
"Bart Simpson" <[email protected]> wrote in message news:<[email protected]>...
> Hi,
>
> We are trying to do a clustered deployment on the WLS 701. I am trying to figure
> out all the things that need to be on for clustering to work successfully. Currently
> I am focused on the EJBs. Here is what I have found:
>
> o For remote SLSBs there is no need to specify home-is-clusterable in weblogic-ejb-jar.xml
> because the default value is true?
> o For remote SLSBs there is no need to specify is-clusterable in weblogic-ejb-jar.xml
> because the default value is true?
> o For Entity Beans is-clusterable doesn't make sense?
> o For local Entity Benas home-is-clusterable doesn't make sense?
> o Are the options for home-is-clusterable and is-cluterable enough or do we have
> to specify -clusterable as extra rmic option under ejbc options?
>
> Thank you so much,
> -Bart Simpson
-
We are using Multicast with Weblogic 9.2 for clustering now but are now migrating to Weblogic 10.3 and want to switch to unicast. The documentation seems kind of hazy for Unicast clustering. I have a few questions.
1)The Unicast Broadcast Channel. What exactly is it and what is the format of one.
2) do I need to create a custom channel some where? I don't want to use the default channel because we have many domains in the same server.
3) Generally has does Unicast work and how is it different from multicast.
Thanks,
Tom.I hope this is some what helpfull:
1) it's the channel used for Unicast traphic, if you leave it empty the default channel (listen address) of the Managed Servers are used
2) You can add channels under the Managed Server at Protocol -> Channels
A channel is connected to a hostname + port so it must be unique. I dont see how it would effect multiple domains on 1 machine since they should use unique host+port combinations as well for every managed server.
3) http://en.wikipedia.org/wiki/Unicast
regards,
Jan -
Apache HTTP proxying for load balancing only to a group of non-clustered WL servers
Hi,
We're running WL Server 6.1 SP 2 on Solaris 2.8.
For the Apache HTTP proxy plugin, if you use the WebLogicCluster http.conf option,
do the WL servers you want to load balance across have to be part of a WebLogic
cluster (if you are prepared to do without failover, as I know it would need to be
a proper WL cluster to replicate session info for failover). Can you load balance
across a group of non-clustered WL servers, and maintain the user session to the
one WL server so that it doesn't switch between servers on alternate requests for
the same user session, or must the servers be configured as a WebLogic cluster?
Paul
We find that if you have a collection of WL servers that are not configured as a
cluster, that it will load balance alternate requests to each server, but it will
not pin a user to a single machine according to their session so for 2 servers, 2
differetn sessions get created, one on each machine.
Is this because it doesn't normally do this, but sends the user alternately to a
primary then secondary which works in a cluster because the session is replicated.
I thought the secondary was only used when the primary failed.
We're running WL Server 6.1 SP 2 on Solaris 2.8. >
> For the Apache HTTP proxy plugin, if you use the WebLogicCluster http.conf
option,
> do the WL servers you want to load balance across have to be part of a
WebLogic
> cluster (if you are prepared to do without failover, as I know it would
need to be
> a proper WL cluster to replicate session info for failover). Can you load
balance
> across a group of non-clustered WL servers, and maintain the user session
to the
> one WL server so that it doesn't switch between servers on alternate
requests for
> the same user session, or must the servers be configured as a WebLogic
cluster?
You don't have to use the clustering option. To get failover, you'll have to
use the JDBC persistence option of WL.
> We find that if you have a collection of WL servers that are not
configured as a
> cluster, that it will load balance alternate requests to each server, but
it will
> not pin a user to a single machine according to their session so for 2
servers, 2
> differetn sessions get created, one on each machine.
>
> Is this because it doesn't normally do this, but sends the user
alternately to a
> primary then secondary which works in a cluster because the session is
replicated.
> I thought the secondary was only used when the primary failed.
The primary/secondary stuff requires clustering. If Apache continues to
"load balance" after the first request, you need to either use JDBC session
persistence or use a different load balancer (like mod_jk for Apache or a
h/w load balancer with support for sticky).
Peace,
Cameron Purdy
Tangosol, Inc.
http://www.tangosol.com/coherence.jsp
Tangosol Coherence: Clustered Replicated Cache for Weblogic
"Paul Hammond" <[email protected]> wrote in message
news:[email protected]...
>
-
How to create Groups and Group Leaders in Clusters.
Hi,
As we know in unicast there is one to one communication and there are groups to control the Thread
Management, How the Groups and the Group Leaders are created.
Regards,
Vardhan.Unicast clustering uses TCP/IP sockets to pass cluster messages between members. To avoid requiring each cluster member
to have connectivity to every other cluster member, WebLogic Server uses a group leader strategy whereby the oldest member
of the group (in other words, the server that was started first) is designated the group leader. All members of the cluster
connect to the group leader so that the group leader acts as the relay point for cluster messages between members.
If the group leader goes down, the next oldest member becomes the new group leader.
As you can imagine, the group leader strategy works well for small groups but becomes less efficient as the number of members
of the group grows large. As such, WebLogic Server uses a multiple group leader strategy where it limits the number of members
in a group to 10. If the cluster is larger than 10 members, WebLogic Server splits into two or more groups, each with their own
group leader. The group leaders themselves are all interconnected to minimize the number of hops that a cluster message must
traverse to reach all cluster members. -
Hi,
We have an application in which we would be having multiple machines. We will
be interfacing with external systems asynchronously. On receiving a message from
these systems, we would want to start back a Workflow that we had put to sleep.
Do we need to cluster all the WL instances to acheive this?.
We would have 2 instances/ machine and 4 such machines.
If the message received is posted to any queue/event , would PI pickup on any
of the instances if we do not "cluster" the Queues.Whether to use shared data store or shared nothing is really dependent on why you are clustering, and whether or not your shared data storage has its own built redunecy or not.
If you are implementing the cluster to ensure high availability and diaster recovery then you need decide whether or not the shared data store is a single point of failure. Usually with shared data store the shared data store is mounted to either something like a SAN or some other remote mounted storage system. If this storage system is itself highly available then you can use the shared data store to reduce total disk usage and still have a highly available environment.
If on the other hand the data storage mechanism isn't highly available then you would want to consider the shared nothing cluster in order to eliminate a single point of failure. Now keep in mind that if just mount both instances data stores to the same SAN then you still have a single point of failure and you may as well go with a shared data store.
No matter what clustering option you select you will still be able to use replication and reverse replication. The replications services implement a cluster aware interface that ensures that they only run on the master node of the cluster. -
In my company we are using 2 publish server for internet and 2 publish server in intranet and one author server.
We have around 10 Intranet sites and 10 internet sites. We are planning to migrate 80 more sites in internet and 100 in intranet.
Content size is approx 20GB in internet and intranet it will increase it to 3 times after migration.
No of authors will be 500+ currently it is 100.
I am thinking of upgrading CQ publish to 64 bit OS and use SAN disk increase memory and cpu .Also business wants to make author env resilient so i am planning to use shared data store clustering in author env and multiple publish server ( no clustering )
My question is will i be able to use replication and reverse replication in this case ? Please let me know your view on this approch.Do you suggest to use shared nothing clustering in author env?
high level architectural of my solution is ( except author cluster with shared data store)
http://dev.day.com/content/docs/en/crx/current/administering/cluster/_jcr_content/par/imag e_0.img.png/1350922018297.png
Regards,
ManishWhether to use shared data store or shared nothing is really dependent on why you are clustering, and whether or not your shared data storage has its own built redunecy or not.
If you are implementing the cluster to ensure high availability and diaster recovery then you need decide whether or not the shared data store is a single point of failure. Usually with shared data store the shared data store is mounted to either something like a SAN or some other remote mounted storage system. If this storage system is itself highly available then you can use the shared data store to reduce total disk usage and still have a highly available environment.
If on the other hand the data storage mechanism isn't highly available then you would want to consider the shared nothing cluster in order to eliminate a single point of failure. Now keep in mind that if just mount both instances data stores to the same SAN then you still have a single point of failure and you may as well go with a shared data store.
No matter what clustering option you select you will still be able to use replication and reverse replication. The replications services implement a cluster aware interface that ensures that they only run on the master node of the cluster. -
OTV using Unicast adjacency does it still need pim and msdp
Hi
We have a working OTV enviroment. I have been asked to created a second overlay interface for a DMZ enviroment
Where currently using
otv adjacency-server unicast-only.
but i still see a number PIM interfaces and we also use the MSDP conficugration which i presumed to be only for the Multicast mode
Can some one please clarify if using unicast mode do i still provision all PIm and msdp confguration to get unicast configuration working
documentation is unclear
the
PIm information is on physical interface between both DC and interconnect between both 7k Switches
MX-N7K2-AGG2# show ip pim ne
PIM Neighbor Status for VRF "default"
Neighbor Interface Uptime Expires DR Bidir- BFD
Priority Capable State
10.44.248.57 Ethernet1/19 8w3d 00:01:29 1 yes n/a
10.44.248.49 port-channel61 36w3d 00:01:39 1 yes n/a
OTV devices as are using igmp ver3
MX-N7K1-OTV1# show ip igmp interface brief
IGMP Interfaces for VRF "default", count: 0
ip msdp originator-id loopback0
ip msdp peer 10.51.0.7 connect-source loopback0Hi,
I am not very sure on how Nexus works related to OTV unicast, but in case of ASR1K you definitely do not need any multicast configuration when bringing up OTV using "adjacency server unicast-only" option.
If you need configuration details related to ASR1K, I have a small blog wherein I have configured the same [both for unicast / multicast OTV cores].
Do have a look if needed: http://stayinginit.blogspot.in/2013/12/basic-unicast-overlay-transport.html -
Clustering vs. parallel server
Pardon me for asking silly questions. I need to clarify the following:
We are going to install the cluster server software for Oracle for solaris, which means we will have clustering option for our db. Currently we install Oracle8.1.7 Enterprise edition as a standalone db server. Does this same db can be used with the clustering software? Do I need to install oracle parallel server?
or is the enterprise edition same with parallel server?
Thanks
MinnyThere is little point in clustering iAS for forms and reports, because the client connection will always crash if one node fails; the user will have to connect again. What I usually do is configure a load balancer (hardware or software) in front of the the two (or N) servers.
-
Clustering Report Server on 10g
Hi,
I have one machine installed infrastructure and Middle Tier. I have created a manual report server and it works fine but when the load is more is take more time. I want to create report servers with clusters option. How do i create cluster report does it need 2 server or it is ok with one server.
RegardsHi Martin,
Well we are trying to run a report without exactly specifying the name of reports server anywhere, e.g. in database or in form or anywhere else. Now if I do not supply a reports server name using RUN_REPORT_OBJECT, the error it displays is FRM-41211: Integration Error : SSL failure ... However if I specify the reports server name in the form, the reports run perfect. Also the name of reports server is specified in rwservlet.properties.
Now the question goes as follows:
Can I run my report from Form without specifying the name of the Reports server anywhere at all. This is so because either an in-process reports server should be picked or the one which is entered in rwservlet.properties should be pickec up by default. Please correct us if we are wrong. once we are through with it, we have to move to Oracle 10gR2 concept of reports server.
Thanks in anticipation,
Ruchi/Saurabh -
WebLogic Express and Clustering
Does weblogic express offer the same clustering options as weblogic server
with regard to HttpSession replication?
I'm looking at an architecture where the web container is separated from the
ejb container (for security reasons) and would like to avoid paying the
higher WLS license.
- Avi
s/\be(\w+)/e-\1/g;
Yes, we just added WLX clustering as a package available. Contact
[email protected] for more information.
Thanks,
Michael
Michael Girdley
Product Manager, WebLogic Server & Express
BEA Systems Inc
"Avi Kivity" <[email protected]> wrote in message
news:[email protected]..
> Does weblogic express offer the same clustering options as weblogic server
> with regard to HttpSession replication?
>
> I'm looking at an architecture where the web container is separated from
the
> ejb container (for security reasons) and would like to avoid paying the
> higher WLS license.
>
>
> - Avi
> --
> s/\be(\w+)/e-\1/g;
>
>
>
-
Which APs allow clustering with an AP541N?
I have one AP541N and need to extend the wireless network with a second wireless AP. Which models of AP can I use with the AP541N to achieve a 'cluster', please? Is it just with a second AP541N or can I use other Cisco APs to achieve the cluster?
Am I correct in thinking that doing this will allow wireless clients to 'roam' seamlessly between the two APs?
Thanks for your help.Hi, My name is Eric Moyers. I am a Network Support Engineer in the Cisco Small Business Support Center. Thank you for using the Cisco Community Post Forums.
To the best of my knowledge, the AP541N will only cluster with other AP541N's.
The clustering option with AP541N's is strictly for ease of management. It allows multiple AP541N's to share configuration information to speed setup, reduce wireless channel interference and simplify management. What is actually shared between the AP's is Wireless Network Identifier, Security features, User Names and Passwords, Traffic Priorities, Radio Settings and Wireless Interface settings.
The Ap541N has what is called Fast Client Roaming capability, not Seamless Roaming. The difference here is that seamless roaming uses a Wireless LAN Controller to help monitor signal strength between the AP and a connecting client. When the signal degrades to a certain point the WLC will automatically switch the wireless client over to the next strongest signal.
With Fast Client Roaming the AP is using a Fast Roaming Protocol. With this there is a slight disconnect and then reconnect from one AP to the next strongest one. Note: you may experience lost packets as the device jumps between AP's, but the connection should not be lost. Should being the main word. Depending on environment and congestion, if the reassociation takes too long, the connection may be dropped.
If you have any other questions please let me know.
Thanks
Eric Moyers
Cisco Network Support Engineer
SBSC Wireless and Surveillance SME
CCNA, CCNA-Wireless
1-866-606-1866
Maybe you are looking for
-
Time Machine doesn't restore my Business Folder in Mail, which is empty after I installed Mavericks. It will restore just the folder but nothing in it. What am I doing wrong? Also, Time Machine doesn't go back more than about a year. I thought it wou
-
Key Operation and Sub operation PM order
Hi, What is the key (table - field) which determines the relationship between the main operation and the sub operation? I mean how can you detemrine that suboperation X is indeed a sub operation of this particular operation Y? You have AUFNR, AUFPL,
-
Update amount when quantity has changed in MIRO
For INVOICE in MIRO tansaction, i supply PO number in PO reference.. When i change the quantity in item level for PO , Amount corrresponding should get updated before saving . Amount = quantity * Price. I have gone through uer exits and BADIS Kindly
-
Wrong delivery status correction
Hellow! I have delivery with good movement status C (compleated ) but no material document attached to the delivery in document flow. Does any report exists to c Andrey Garshin.
-
ODI: Is ODI 10g compatible with SOA 11g ?
hi all, I have a requirement where ODI should invoke web services in SOA 11g fusion middleware. Is it required that i use ODI 11g? Or is ODI 10g compatible working with SOA 11g. In case if i have to use ODI 10g to invoke webservices in SOA 11g den do