Unable to deploy Large Size file

Hi Team,

I am unable to deploy a large size file through opendeploy to target servers.

At first time when we received the issue we found that parameter blockMaxWaitTime is set to 1800 seconds (30 Minutes).

We assumed that as the files of size big size (115 MB), so by the time the file transfer is completed on the servers, it took more than 30 minutes and deployment are getting failed.

By working on above assumption, we reset the parameter blockMaxWaitTime to 0 (Zero). By setting the blockMaxWaitTime="0" the webservers will wait indefinitely and wont timeout after 30 minutes.

But even after this change, deployment is still failing. Other deployment are working fine so issue happens only in case of deployment of Large size File.

We are doing SSL based deploymnets.

Please advise about it.
Thanks in advance.

Vivek
ODR : 731

Find more posts tagged with

Comments

Adam Stoller

Some questions:

before responding...

VivekPolaris

Thank you very much for your reply. Plesae see my inlin response for your queries.

Is this a transactional or non-transactional deployment?
>> Transactional.

Are there errors in the log files (if so, which log files and what errors) or does the deployment just seem to hang?
>> it does not hang. Its coming in deploymentspecic logfiles stored at Od-home/logs folder. Unable to connect my office network so will provie error logs somewhat later.

Are the Receivers running on Windows or Unix?
>> Linux.

Have you checked the amount of space availabl1e in the target path of the Receiver?
>> Yes checked. enough Space is available.

Please let me know if more info is required.

Thanks again!!

VivekPolaris

Below Error messages are being loogged.
Please find the attached deployment log for the reference also.

LIB 00024: 2014-04-18 12:47:32 ERROR: Could not send phase start message.
LIB: 2014-04-18 12:47:32 Receiver-ERROR: DYi^H.
LIB: 2014-04-18 12:47:32 ERROR: Failed to deploy local_directory[.] of area[/iwmnt/international/main/CITI/japan/jpgcb/WORKAREA/work/webserverdocs] for filelist[/international/main/CITI/japan/jpgc
b/WORKAREA/work/filelist/deployment/japan/cjp_CBOL_authoring_web_uat_filelist_1855656]
LIB: 2014-04-18 12:47:32 ERROR: Failed 1 of 1 filelists

ENG: Thread-6244 After COMMIT, res=-1
ENG: Thread-6244 exiting commit
ENG: Thread-6244 entering awaiting-rollback
ENG: Thread-6244 still waiting for 2
ENG: Result after commit -1
ENG: Thread-6244 exiting awaiting-rollback
ENG: Thread-6244 entering rollback
LIB 00024: 2014-04-18 12:47:44 ERROR: Could not send phase start message.
LIB: 2014-04-18 12:47:44 Receiver-ERROR: a^?U^TNi^Hxa^?Uo^L9^F^TNi^H a^?U.
LIB: 2014-04-18 12:47:44 ERROR: Failed to deploy local_directory[.] of area [/iwmnt/international/main/CITI/japan/jpgcb/WORKAREA/work/webserverdocs] for filelist[/international/main/CITI/japan/jpgc
b/WORKAREA/work/filelist/deployment/japan/cjp_CBOL_authoring_web_uat_filelist_1855656]
LIB: 2014-04-18 12:47:44 ERROR: Failed 1 of 1 filelists
ENG: Thread-6244 After ROLLBACK, res=-1
ENG: Thread-6244 exiting rollback
ENG: Thread-6244 entering awaiting-cleanup
ENG: Thread-6244 still waiting for 6
ENG: previousPhaseStr:rollback
ENG: currentPhaseStr:awaiting-cleanup
ENG: Thread-6244 exiting awaiting-cleanup
ENG: Thread-6244 entering cleanup
LIB 00019: 2014-04-18 12:47:45 ERROR: Could not read content size from remote.
LIB 00022: 2014-04-18 12:47:45 ERROR: Could not read server status, client OK.

Adam Stoller

[...]transactional, Linux, not-hanging, space available[...]

When you say you've verified space is available, are you verifying that there is at least three-times the amount of space needed for the individual file (in this case, 3 * 115 MB = 345 MB) available on the target paths' partition (/ujfpgcb2/docs/jpgcb_uat)?
Could you show us the output from the following command run on all target systems (I cannot tell from what you posted whether this is a single point-to-point deployment or a fan-out deployment)

% df -k /ujfpgcb2/docs/jpgcb_uat

There clearly seems to be some non-ascii characters involved in the log files (presumably Japanese characters) - I have no idea whether that's an issue here or not. I'll assume "not" for now, but you may have to revisit that later.

If this is part of a fan-out deployment - we'd need to see the sender and receiver micro logs for all hosts in the replicationFarm.

VivekPolaris

Thank you very much Ghoti for your reply.

1. I checked the result of commnad on all target servers. The result is almost same and indicates that enough space is there on all Servers:

df -h /ujfpgcb2/docs/jpgcb_uat
Filesystem Size Used Avail Use% Mounted on
/dev/mapper/gcbvg-ujfpgcb2lv
20G 12G 7.7G 59% /ujfpgcb2

2. we are working for the client from Japan, so there can be Japanese chars in file deployed.

3. This seems a filelist deployment. I cheked the deploy config file and found below:


<sourceFilesystem area="$srcarea3^"  filelist="$filelist3^">

In logs also, a list is getting created, which has the file names with path to be deployed.

For quick reference, I am attaching deploy cofig file and Od BASE logs for All Servers for the failed deployment.

Thanks again!!

Adam Stoller

This is definitely a very big and complex deployment issue. You'd probably be wise to get HP Support involved.

Part of the complexity here is:

What I would try to do is isolate and minimize the problem as much as possible. Using a file list that only contains the file you think is having problems getting deployed, try deploying it to each one of the receivers one-by-one (instead of as a fan-out to all of them at once). Try anything else you can think of to reduce the number of variables involved to the point where you can get a reliably reproducible issue - that will make it much easier to work with Support (and anyone else for that matter).I don't really have much else to offer - sorry.