ESA deployment in an ISP environment

Hi All,
I am to deloy an ESA in for an ISP and some questions came up since i mostly deploy in enterprises/offices with a firewall (DMZ) which is very straight forward. For an ISP however, i am having trouble understanding how mail flows from their clients to the ISP then to the internet and back. Can anyone explain how this would be done:
1. the ISP want mail from their clients (who host their own MTAs) to be passed through spam filters (hosted by the ISP) before going out to the internet to avoid the client IPs being blacklisted all the time due to sending spam.
2. ISP wants incoming mail for their clients to go through the ESA before being forwarded to the clients MTAs, which as i said before, are hosted in the clients LAN.
Questions:
1. Where in the ISP network will the ESA sit (is it in a LAN just behind the core router or in the ISP internal LAN behind a firewall?)
2. Does the ISP have to do some kind of re-direction for SMTP traffic to pass through the Ironport ESA before going to the clients (for incoming mail)
3. Does the client and/or ISP need to change their MX records.
I would reaally appreciate a breakdown of how this deployment would be done.
Thanks.

Henry,
I would also recommend that MX records be changed to allow inbound smtp traffic to pass through your boxes. In the absence of a hardware load balancer you can use two equally weighted MX records, one pointing to each box.
On the inbound side you will end up with recipient validation issues. Spammers and others can and WILL send messages to a lot of bad addresses. If you can't validate them at the edge on the way in then you will get stuck with them on your appliances and they can clog up queues if not managed carefully.
Your options for outbound are limited:
1) Transparent proxy in "stealth" mode
A transparent proxy that would capture SMTP to any address is not a function of an Ironport ESA but you could potentiallty set one up that forwarded everything to the Ironport appliances.
A transparent proxy gets complicated in an ISP environment. Customers may not be so happy with "hidden" traffic manipulation. You would have to turn off Received headers which is not RFC-compliant for edge systems and eventually customers would figure out that the IP where other systems see the mail coming from is not correct. This would make troubleshooting rather complicated. Not to mention that it is a bit heavy-handed.
2) Transparent proxy with headers getting added
3) Block direct outbound so they will be forced to relay through you
Many ISPs do this for residential customers but it tends to be less commonly applied to business customers running their own MTA.
4) Make relaying optional
If customers are going to be looking to you for help with blacklist and delivery issues then optional might not be the best method unless you SWIP networks to your customers (non-portable) so they become the contact of record.
Be aware that MS Exchange can spit out a lot of cruft that will add to your load. Things like Out-Of-Office replies and bounces from email sent to bad addresses can be a measureable percentage of the total email traffic. You can even see things like bounces including a copy of original attachments (WHY?). Just watch your undelivered recipients so that it doesn't get too high.

Similar Messages

Create a new portlet in Deployed Portal in production environment

are there any ways to import or create a new portlet in Deployed Portal in production environment?
Edited by: user8322798 on May 1, 2011 7:26 AM

This can done via WSRP proxy portlets and streaming desktops. First, you'll need to have a WSRP producer setup somewhere. This could be another WLP webapp with portlets, or another server altogether. Or, you can use the JSR 286 WSRP import tool from within the Portal Administration Console (I think it's under Services | WSRP | Import Tool) -- this will allow you upload .war(s) of JSR 168 or 286 portlets, which will be turned into WSRP producer(s).
Then, you can use the Portal Administration Console to register a WSRP Producer, and then add portlets from the producer to your desktop (http://download.oracle.com/docs/cd/E15919_01/wlp.1032/e14235/chap_fed_books_pages.htm#FPGWP690). Additionally, once the producer has been registered in Portal Adminstration Console, an adminstrator user can use the Dynamic Visitor Tools from within the streaming desktop itself to add wsrp proxy portlets to the desktop (http://download.oracle.com/docs/cd/E15919_01/wlp.1032/e14243/dvt.htm#PDGWP691).
It is not possible to add new local .portlet files to a deployed application in production mode. This requires adding the file artifacts to the .ear/.war and redeploy the application.
Greg

Workflow Custom Activity deploy in multi server environment

I have been working on a project that involves developing a custom workflow activity for SharePoint 2013. I am developing it in a single server environment working with http.
My problem occurs when deploying to multi-server environment with https (WFE, APP). My question is how to deploy my solution to the new environment.
The steps:
Create a project - c# activity library
Add a workflow activity, add .xml
Deploy the .dll and .xml of project to:
"C:\Program Files\Workflow%20Manager\1.0\Workflow\Artifacts" "C:\Program Files\Workflow Manager\1.0\Workflow\WFWebRoot\bin"
net sotp "Workflow Manager Backend"
net start "Workflow Manager Backend"
Deploy .DLL to GAC
- Created MSI using install shield in VS2010 to add .DLL to GAC
- Verify .DLL in GAC by going to c:\windows\assembly and %windir%\Microsoft.NET\assembly
iisrest
Deploy WSP to SharePoint, activate feature, open SharePoint Designer 2013 and choose the custom action that now appears when creating a 2013 workflow
To recap we have workflow manager on the APP server and and the workflow client on the WFE. We deployed the .DLL and .XML to the workflow manager (APP) only. The .DLL is deployed
to/in the GAC on the WFE and the APP. We are able to see and create the activity in Designer 2013 and we deploy the workflow to a normal SharePoint list. When we run the workflow we do not get any errors in the ULS logs, event viewer or Workflow Manager Debug
Logs (event viewer also). The site is not created though. We believe the issue is that the custom C# (.DLL) is not being ran.
This all works fine and dandy on my single server environment. Workflow is working like a charm. How can we trouble shoot what the issue is if we are not finding any errors?
Is there a step that we missed or some other place we need to look for logs? Would the ULS logs show the site creation or show running our custom code? Currently it does not show anything when we run the workflow.
Let me know if this is unclear or if anyone needs more information. Thanks

Hi,
Here is a workaround for your reference:
We can develop a custom WCF service instead of the Custom Activity in SharePoint. And then use the service from workflow. It use a separate dedicated server for workflow without having any reference to SharePoint DLLs from inside workflow.
Here is a similar thread for your reference:
https://social.technet.microsoft.com/Forums/systemcenter/en-US/d462ca07-9861-4133-948a-fc9771306cb1/custom-workflow-how-to-go-from-single-server-to-multiple?forum=sharepointdevelopment
Thanks,
Dennis Guo
TechNet Community Support
Please remember to mark the replies as answers if they help and unmark them if they provide no help. If you have feedback for TechNet Subscriber Support, contact
[email protected]
Dennis Guo
TechNet Community Support

SAP ME SDK - WebServices Deployment to a Production Environment

Hello all,
I was wondering if anyone else out there was having the same problems that I was having while developing a Web Service through the SDK Eclipse environment?
I followed the tutorials within the "SDK Training.pdf";; that came with taking the SDK class and was able to create the Web Service (page 264) without any problems on my development server. Deployed through the Eclipse environment. However when performing an actual build to be deployed on a Production server the Web Services were not carried over cleanly.
I did manage to get it all working after a lot of trial and error as well as creating a new directory structure to place the wsdl files in, which ironically is the same directory structure on the development server (a good thing that I have a D: on both the dev and prod server). Not a very pretty situation. It should come all nice and neat for a production build similar to deploying the SAPME.sca file.
I was just wondering if anyone else in the community had the same problems that I did?
Also, was wondering if there was any future enhancements to the SDK to make this process more streamlined?
Thanks,
Mike

Hi Troy,
I have already done exactly as you suggested, which is also documented in the training information for a production build. I am able to get all of my custom code/activities/jsp pages working within ME without any issues. However, I was not able to get to my web services through the JSP pages that I wrote. Following the example in the Training Documentation given to us from the SAP Trainer.
To get this to work I had to take the <vendor>Web.sda.ear file and move that to my Production server to install and run. That is the part that I was dicussing it not being a clean install.
I will take another look at the SDK Implementation Guide, however the Training Doc has examples, which many people, including myself learn best from.
My question was really has anyone else in the community run into this same issue? As well as, are there any improvements?
You have answered the second question with regards to ME 6.1. I appreciate and thank you for that.
I am looking forward to the ME 6.1 ramp up and release.
Thanks,
Mike

How to deploy reports from one environment to another in CoD

Hi Experts,
I'm actually in training with an Oracle partner in order to learn how to create reports in CoD.
It seems quite easy.
But the trainer here told us to develop directly the new reports on the Prod environment.
I've asked why, cause we also have Dev and Test environment.
And he answerer me that there's no way to deploy reports from one environment to another, the only way is to redevelop it on each environment.
As for me it seems impossible not to be able to deploy reports from dev to test and then test to prod, I'm directly asking experts.
So could you please confirm this or not, and maybe provide me an example of what you're doing on your side to copy from one environment to another.
Thank you in advance.
Brice

H,
reports cannot be migrated at this time, indeed.

Cisco ESA Deployment Question

Hello,
I have a question about the ESA deployment, In case I have 1 ESA deployed in my network with MX record and the public IP is natted to the ESA IP located on the DMZ.
If the ESA goes down for any reason like power failure, still I can recieve and send email or not? I mean in this case, can the device work in fail-open mode to relay the emals but without applying the policy?
Or in this case, the mail system will be completly down, and I need to add 2nd MX record (High avaailbility) to ensure that the email system is UP?
Thanks.
Ahmad.

This is very logical for me.
I had a discussion with Cisco SE and he insists that ESA acts as a proxy not as email server, so the mail server would still deliver email but without any ESA policies applied to it.
Here, if you point the exchange server to the ESA and the ESA is down, then the recieving and sending will be down, and only the internal emails will be working.
Thanks.
Ahmad.

Error while deploying Collections to Dev Environment

Hi
I am trying to deploy DC Modules (Collections) from OWB Repository to DEV1 environment.
I am doing it as follows.
OMB+> {color:#0000ff}OMBCONNECT user/pwd@host:port:schema
{color}Connected.
OMB+> {color:#0000ff}OMBCC '/DQ'
{color}Context changed.
OMB+> {color:#0000ff}OMBCONNECT CONTROL_CENTER user/pwd@host:port:erpdev1 USE REPOSITORY 'repository'
{color}Control Center connected.
OMB+> {color:#0000ff}OMBCREATE TRANSIENT DEPLOYMENT_ACTION_PLAN 'MY_DEPLOY_PLAN'
{color}Action plan MY_DEPLOY_PLAN created.
OMB+>{color:#0000ff} OMBALTER DEPLOYMENT_ACTION_PLAN 'MY_DEPLOY_PLAN' ADD ACTION 'MY_COLL_PLAN' SET PROPERTIES (OPERATION) VALUES ('REPLACE') SET REFERENCE COLLECTION 'DC_INV_CUSTOMER_ITEMS'
{color}Action plan MY_DEPLOY_PLAN altered.
OMB+> {color:#0000ff}OMBDEPLOY DEPLOYMENT_ACTION_PLAN 'MY_DEPLOY_PLAN'
{color}{color:#ff0000}OMB05619: No Location defined for Module
{color}
Can you please let me know if I am missing something.
What is this Location and where do I need to define this ?
As per my understanding, the Collections are logical groupings of many objects defined in a module called DC_STAGING.
The Collection is already created in the OWB. How can I specify the location of this in the commands/script.
Appreciate any help.
Thanks and Regards
Sameer

Hi Dallan
As I mentioned earlier the Collection is a logical grouping of Tables and Mappings from one of the Module 'DC_STAGING'.
So I am getting the tables and mappings references from the collection as follows.
OMBRETRIEVE COLLECTION 'DC_INV_CUSTOMER_ITEMS' GET TABLE REFERENCES
So I think we have to deploy all the tables and mappings in the Collection and deploy them as usual.
Thanks and Regards
Sameer

Windows 8 App Deployment in School Lab environment?

After doing some reading and research on Windows 8 Store app deployment from sources like
http://www.microsoft.com/en-us/download/details.aspx?id=39685 and this forum and I am getting pretty depressed. Here is one part of my scenario which is pretty simple I think.
Lab of 20 Dell 3011 AIO Touchscreens that were depoyed with Win8Ent via MDT2012U1. For simplicity I will talk about a subset of my problem. I have 3 classes of Kindergartners who use the lab. The students aren't capable of really logging
in themselves so the teacher logs them all in to a shared account for each class since all they do is web / app stuff. I don't want to create 60 accounts for this young of kids since they can't handle the process anyway.
Here is my understanding of what I would have to do if I wanted to install Fresh Paint (a great free app) for all the students to use.
1. Create a minimum of 4 Microsoft accounts due to device limits of store app install.
2. Assign / attach those Microsoft accounts to specific lab machines (while logged in as each shared student account).
3. Login to each of my 20 lab machines as each of the 3 share accounts (which can't have roaming profile since Windows <-> Microsoft account association now needs to be unique). This would be a minimum of 60 individual logins for this one
app and one grade.
4. Install the app on a per user / per device basis.
5. Potentially repeat process any time an update is released since the students can't handle it themselves.
I think this is the majority of the process, am I leaving any important steps or details out? This sounds like it would be HOURS of work to install ONE app for ONE grade. I could potentially need 5-10 unique apps per grade.
This just isn't feasible on any type of scale. Even this limited amount of devices / users is overkill when dealing with users who are to young to maintain their own devices or exist in lab environment. In reading the doc refrenced above it doesn't
seem like there is a solution for this offered by MDT, SCCM or Intune? Please tell me I am missing the magic deploy button? Also I don't even want to begin to think about apps that aren't free and managing that.
Thanks for any helpful info.
Brian Hoyt

In reading the doc refrenced above it doesn't seem like there is a solution for this offered by MDT, SCCM or Intune? Please tell me I am missing the magic deploy button?
No, you're not missing the magic deploy button :(
If a Windows Store App (aka Modern UI app aka Metro App) is offered via the Windows Store, and not offered directly by the application developer, the only legitimate way to get that app is via the Store, and that requires an MS account.
ConfigMgr offers two "deployment" methods;
a) deep-link into the web-based Windows Store. The logged-on user must have an MS account and use that account to authenticate to the store. even if the app is free.
b) sideload. requires the appx installation package.
There are some other suggestions, but they are unsupported, and rather complicated:
http://social.technet.microsoft.com/Forums/windows/en-US/763dc3d6-8dd2-4157-8f49-159f0826e97a/install-windows-store-apps-to-all-users-or-where-to-get-the-appx
http://social.technet.microsoft.com/Forums/windows/en-US/04dd1938-a558-4cbe-ba70-b89d2d1eb95c/windows-apps-offline-download-and-install
http://social.technet.microsoft.com/Forums/windows/en-US/72ccc40d-76aa-4325-92ae-27386b424e0f/sideloading-applications#bddbcd7f-fec3-4132-b573-b5e3d1aef57f
http://social.technet.microsoft.com/Forums/windows/en-US/b7574287-d134-4a74-b227-f3dd793ab46b/download-appx-package#64745402-2254-4760-bc1f-f7483fc49d35
The last time I spoke with our MS contacts (deployment engineers @ MS), even those guys can't get the MS free apps to sideload.
All the documentation around sideloading appears to be catering for situations where a company has developed that app themselves, or, has commissioned a developer to create a custom app for the company.
Apps which are offered via the Store, seem to only be available via the store. I'm not sure if there is some kind of contractual obligation with MS which demands this? (i.e. in the store means no distribution via other means?)
At this time there are no options to federate an on-premises AD to the Store either. Which would be a nice option, even if it limited the user to free apps.
Don
(Please take a moment to "Vote as Helpful" and/or "Mark as Answer", where applicable.
This helps the community, keeps the forums tidy, and recognises useful contributions. Thanks!)

Error Deploying Mapping in Production Environment

All,
I have 3 locations: APPS (source,) STAGING (target,) and SALES_DM (target.)
I am trying to deploy a mapping, via deployment manager, to my PRODUCTION runtime environment. The mapping is part of the STAGING schema, has two source tables from the STAGING schema, and after a simple join performs a lookup to APPS, STAGING, and SALES_DM tables, and eventually loads to another table in the STAGING schema after various transformations (it is a complicated mapping that deals with invoices and orders for a Sales DataMart.)
I have deployed and executed this exact mapping in several test instances. I have also deployed and executed many other mappings in the STAGING schema in PRODUCTION. However, I am encountering an error when I try to deploy this specific mapping: PL/SQL: ORA-00942: table or view does not exist.
This error occurs 4 times (there are 4 warnings with different line numbers.) When I go to the lines in the code (using the line number + line number where the / in the main is found trick) the same STAGING table is in every line. Since this mapping executes as STAGING, it is a little weird that the error message is indicating a permission issue with one of its own tables. Is it possible that this specific table is not the real issue?
Here is the main difference I can see between my PROD and TEST environments. In TEST, the APPS location is registered to APPS, where as in PROD it is registered to STAGING. However, in both instances I am positive that the appropriate grants on source APPS tables have been given to STAGING, and in PROD I have created synonyms for the source APPS tables in STAGING (i.e. for the Applications order headers table I have created the synonym oe_order_headers_all.)
I have had to do a lot of troubleshooting with apparent permission issues before, but am completely baffled on this one.
Any thoughts or ideas are much appreciated.
Regards,
Ryan

Hello,
Since you have one PRODUCTION runtime, the targets STAGING and SALES_DM runs in the same instance. If you register both locations properly in Deployment Manager, OWB will figure out that you deploy to the same instance and will use schema.table naming in your mapping PL/SQL package for efficiency, not any DB link.
Because of this you/SALES_DM must (manually) GRANT SELECT on SALES_DM.lookup_table TO STAGING in the PRODUCTION instance (you must grant directly to the STAGING user, not to any role held by the user).
I presume you have checked that your lookup tables have been deployed, as reminded by another forum participant earlier...
Regards, Hans Henrik

How to manage a debug deployment in a production environment

One of the advantages of the Base Deployment is that the test station is locked down, for example operators cannot open and edit LabVIEW VIs. AFAIK a debug test station is wide open like a development machine, except for TestStand User Manager restrictions.
How can I have a test station that is locked down, and then open it up for debug as needed? Dual booting the computer seems like a solution. Or maybe just have an open debug station and trust the trained operators to not mess with it. Can anyone share their solutions? Thanks.
cc

Hi,
You make a good point about TestStand editing. That really leaves only the issue of LabVIEW development environment allowing uncontrolled editing of VIs.
Personally I'm good debugging with base deployment, however I have worked with many companies who want full debug capability.
Here’s my refined questions,
1. Can I create one deployment image that works both with TestStand Base Deployment license and Debug license? Do I have to add anything special other than the Sequence Editor to get TestStand to run as debug?
2. If I install LabVIEW development environment onto the same machine as the deployment image in 1, will I be able to step into and edit LabVIEW VIs? (As long as I set the LabVIEW adapter to development.)
Then the debug station would be locked down with TestStand User Manager and LabVIEW adapter set to runtime. And it would have the same TestStand and LabVIEW and support files as all the base deployment machines. Does that sound logical to you? Thanks
cc

SharePoint Online - Provider hosted App - Deployed in on-premise environment

Hi
We are evaluating SharePoint online and wanted to know if the following is possible? If yes, can anyone share his/her experience and issues faced?
We want to deploy a provider-hosted app to a on-premise environment. This app will need to connect to SharePoint online. What are the requirements from hosting perspective for the on-premise server? Does the server require anything specific so that it is
able to connect to SPOnline? Should it be outside the corporate firewall?
What is the difference between SharePoint online standard and dedicated plans, in the context of Provider hosted app. Are there limitations with respect to what can & cannot be done in a Provider hosted app in these environments?
How do I know if am using a SharePoint online standard or a dedicated plan? Where to look for?
I have seen few articles where the provider hosted app is deployed to windows azure and is connecting to SharePoint online. However this is not the case with us (i.e. we do not have windows azure for now)
Regards,
Deven

Below links answer most of your questions:
App- code
http://msdn.microsoft.com/en-us/library/office/fp142381(v=office.15).aspx
How to create high trust apps- http://msdn.microsoft.com/en-us/library/office/fp179901(v=office.15).aspx
How to deploy- http://msdn.microsoft.com/en-us/library/office/jj860570(v=office.15).aspx
Go to billing and subscriptions and you will see what plan you have. Dedicated plans names and options are in this link- http://technet.microsoft.com/en-us/library/dn151715.aspx
Regards
Libin
Whenever you see a reply and if you think is helpful, click "Vote As Helpful"! And whenever you see a reply being an answer to the question of the thread, click "Mark As Answer Feel free to unmark answer if does not resolves your problem.

Cisco ESA Deployment

Hello Everyone,
I have seen deployment guides of Cisco ESA stating that the ESA is to be deployed in the DMZ network
Can We deploy Cisco ESA and also the internal mail server in the same network ?

You can, but you need to make sure that email (inbound and outbound) flows through the ESA and nothing is direct to your email server. Putting ESA on the edge does make the deployment and future troubleshooting easier.
Hope it helps.

Best practice: Deployment plan for cluster environment

Hi All,
I want to know, which way is the best practice for preparing and deploying new configuration for WLS-cluster environment. How can I plan a simultan deployment of ALL of nodes, with out single point of failure?
Regards,
Moh

Hi All,
I get the Answer as followed:
When you deploy an application OR redeploy an application, the deployment is initiated from the Admin Server and it it initiated on all targets (managed servers in the cluster) at the same time based on targets (which is expected to be cluster).
We recommend that applications should be targeted to a cluster instead of individual servers whenever a cluster configuration is available.
So, as long as you target the application to the cluster, the admin server will initiate the deployment on all the servers in a cluster at the same type, so application is in sync on all servers.
Hope that answers your queries. If not, please let me know what exactly you mean by synchronization.
Regards,
Moh

How to deploy jsp in R12 environment?

Hi, I want to write a jsp and deploy in R12 environment. I try to put a simple jsp in the $OA_HTML and restart Apache, but it doesn't work.
Thanks.
Devon

As far as i know its similar to 11i, enable log in jserv.properties and see log in
$APACHE_TOP/Jserv/logs/jserv.log
Consult ur dba.
--Mukul

Deploy EJBs in cluster environment

          I have cluster environment using three levels of weblogic.properties files (global/per cluster/per server). When I put all EJBs into per server properties file in all servers in cluster, I got conflictHandler message for some of the EJBs when bring up the second server in the cluster.
          1. Where the ejb should go (set in weblogic.ejb.deploy property)?
          2. Can I move those conflicted EJBs per cluster or golbal weblogic.properties file to resolve the conflict problem?
          3. If there are weblogic.ejb.deploy entries in both per-cluster properties file and per server cluster properties file, do the setting in per-cluster be overwritten by the one in per-server?
          Thanks in advance.
          key


          Just found out that those beans have been explicitly disable home-is-clusterable. Thanks for all of your help.
          "Vivek Raikanti" <[email protected]> wrote:
          >But I thought, that is the default setting in the weblogic-ejb-jar.xml file
          >?
          >( enable-clustering is true by default ... correct ?)
          >
          >
          > -Vivek
          >
          >
          >Wei Guan <[email protected]> wrote in message
          >news:[email protected]...
          >> In your EJB deployement descriptors, did you enable clustering for your
          >> EJBs?
          >> --
          >> Hope it helps.
          >>
          >> Cheers - Wei
          >>
          >>
          >> "Key Zhang" <[email protected]> wrote in message
          >> news:[email protected]...
          >> >
          >> > I have cluster environment using three levels of weblogic.properties
          >files
          >> (global/per cluster/per server). When I put all EJBs into per server
          >> properties file in all servers in cluster, I got conflictHandler message
          >for
          >> some of the EJBs when bring up the second server in the cluster.
          >> >
          >> > 1. Where the ejb should go (set in weblogic.ejb.deploy property)?
          >> > 2. Can I move those conflicted EJBs per cluster or golbal
          >> weblogic.properties file to resolve the conflict problem?
          >> > 3. If there are weblogic.ejb.deploy entries in both per-cluster
          >properties
          >> file and per server cluster properties file, do the setting in per-cluster
          >> be overwritten by the one in per-server?
          >> >
          >> > Thanks in advance.
          >> > key
          >>
          >>
          >
          >

ESA deployment in an ISP environment

Similar Messages

Maybe you are looking for