To create Opendeploy custom logs.

Hi All,

Jst want to get idea for "how to create OpenDeploy" custom logs for situation for ex, While deploying content to another server how would i get to know that this file haven't get deploy. Any suggestion will b helpful.

Thanks

Manissh Khaparde.

Find more posts tagged with

Comments

David Smith

OpenDeploy already gives you several useful logs out-of-the-box for monitoring and debugging deployments. Both on the source side as well as the target side of the deployment. There's really no need for custom deployment logs. I suggest you read the OpenDeploy documentation. While the error messages can be fairly useless sometimes when something goes wrong, between the messaging on the source side combined with the messaging on the receiver side (which is usually more helpful), you should be able to debug what went wrong fairly easily.

Pauliw

Hello Manish,

Every deployment which happens from the base server, creates its specific deployment log file under <ODNG>/log, this log file will have the status of the deployment, For instance,

1) Type of deployment , the number of files deployed.

2) Error -> then the reason for the error

3) Success -> then good.

In Base server the log would be in the below format

src.<deployment configuration name>.<Base server>.to.<reviver node>.log

The receiver too has the log and its specific detail , the logs in receiver would be rcv.<deployment configuration name>.<Base server>.to.<reviver node>.log

The best practise is to have a folder and have the deployment conf inside that folder. So by this when you call the deployment,OD will create the create the same folder under log and will place the respective log, which would be easy for us to look and debug from them,

Thanks

Paul

David Smith

Pauliw wrote:

The best practise is to have a folder and have the deployment conf inside that folder. So by this when you call the deployment,OD will create the create the same folder under log and will place the respective log, which would be easy for us to look and debug from them,

Why? How is that remotely a "best practice"? The way the logs are named makes doing this redundant.

Pauliw

Its true that logs are named properly under logs folder, when we place deployment configuration in specific directory for a reason, so does the log generation for those deployment gets into the directory under log.

Placing the logs in different directory like within each specific deployment category folder, makes it easier to look for troubleshooting, for any rookie in Opendeploy.

Probably thats why we have lscs and LiveSite deployment configuration in specific folders thus bringing the logs under them.

David Smith

Pauliw wrote:

Placing the logs in different directory like within each specific deployment category folder, makes it easier to look for troubleshooting, for any rookie in Opendeploy.

I disagree - all it seems to do is create extra layers of nonsense, which is unnecessary. A better solution to distinguish individual deployment logs (if that's actually something that would be helpful) might be to use instance values to separate logs out for individual execution of same-named deployments. Be careful when you use terms like "best practice" - that can be deceiving, certainly to a "rookie". Do you work for HP?

Pauliw

No , i do not work for HP

Adam Stoller

Smitty97007 wrote:
Pauliw wrote:

Placing the logs in different directory like within each specific deployment category folder, makes it easier to look for troubleshooting, for any rookie in Opendeploy.
I disagree - all it seems to do is create extra layers of nonsense, which is unnecessary. A better solution to distinguish individual deployment logs (if that's actually something that would be helpful) might be to use instance values to separate logs out for individual execution of same-named deployments. Be careful when you use terms like "best practice" - that can be deceiving, certainly to a "rookie". Do you work for HP?

I think best-practice would be a combination of the two - I haven't used the subdirectory concept that much, but... if you have a multiple oft-used deployment configurations, especially if they can be "categorized" by purpose (e.g.: internet versus intranet), then using subdirectories for the configuration files which will result in subdirectories for the log files is a very reasonable practice. However, the use of distinct instance parameters (e.g. jobid) is also best-practice because without it you cannot run the same deployment for two (or more) invokations concurrently.

Where the subdirectory concept comes into significant help is when there are a lot of distinct deployment configurations which are used constantly throughout the day and little-to-no log file archive / clean-up processes created. Consider the difference between searching through hundreds of log files in one directory versus searching through tens of log files in multiple directories (and then consider the same issue when you multiply the number of log files by a factor or two)

If all you have is one or two oft-used configuration files, then separating them into different subdirectories won't really provide that much impact - but it won't hurt either.

None of this, however, has anything to do with creating custom log files, as was in the original posting.

Manissh007

Hi Paul,

Thanks for the reply Paul. I'll follow the guidelines u have mentioned. Man Happy

Thanks

Manissh Khaparde.

Manissh007

Hi Smitty, Gothi

Thanks all for Urs valuable suggestions. I'll remember the guidelines.

Thanks

Manissh Khaparde.