留言板
Performance
Hello everybody,
In my company, the main developer of a Liferay application has been gone a months ago leaving no documentation about the application. Now, there is an strange issue between liferay and alfresco and I would like to ask you for some new ideas, because we are completely stucked in it. We use Liferay 6.1.30GA3 EE and Alfresco Community 4.2.0 and this is our portal-ext.properties (section regarding only DL config):
session.store.password=true
company.security.auth.type=screenName
dl.store.impl=com.liferay.portlet.documentlibrary.store.CMISStore
dl.store.cmis.credentials.username=XXXXXXXX
dl.store.cmis.credentials.password=XXXXXXXX
dl.store.cmis.repository.url=http://XXX.XXX.XXX.XXX:8080/alfresco/cmisatom
dl.store.cmis.system.root.dir=XXXX (this value is setted to company's name)
session.enable.url.with.session.id=true
I would like to ask you if anybody recongnizes some of these symptoms, and if so, what would be your recommendation.
When we access Alfresco from Alfresco app, all goes right. But when we access documents stored in Alfresco through DocumentLibrary things happen really slow. For example, a for loop iterating over whatever 10 documents takes at least 30 or 40 seconds. It's the same if you try with the first or last 10 documents or else, anyway to obtain a single doc the system takes sometimes until ten or more seconds.
Another thing I found is the alfresco access log files: the same request is made along the day until 20.000 times. There are about 800 repeated requests every 24 hours, many of them for the second of third time, but these 20.000 requests, 10000, 9000, ..., are strange for me (Log files are taking about 80 megabytes/24 hour). AFAIK, there are no changes in the Alfresco config besides DB and root.dir values. In addition, adjusting VM memory parameters to change eden, survivor and old space sizes has not had any apparent effect nor behaviour change.
If we use the migration tool in control panel, the system runs for four or five days until VM crashes (OutOfMemoryError, we have adjusted VM memory parameters again without changes).
Processor use of Alfesco machine is many times at 100% (in the production system), but it's the same. It always takes too much time to deliver a document, even with no CPU actvity. No memory problems, no O.S. swapping. Strangely, using Alfesco application all goes as fast as it should.
There are a lot of hooks and portlets, and in some of them the use of DocumentLibrary contains some harcodes and other strange things.
If anybody recognizes some of this, please, any new idea will be appreciated.
Thank you in advance.
In my company, the main developer of a Liferay application has been gone a months ago leaving no documentation about the application. Now, there is an strange issue between liferay and alfresco and I would like to ask you for some new ideas, because we are completely stucked in it. We use Liferay 6.1.30GA3 EE and Alfresco Community 4.2.0 and this is our portal-ext.properties (section regarding only DL config):
session.store.password=true
company.security.auth.type=screenName
dl.store.impl=com.liferay.portlet.documentlibrary.store.CMISStore
dl.store.cmis.credentials.username=XXXXXXXX
dl.store.cmis.credentials.password=XXXXXXXX
dl.store.cmis.repository.url=http://XXX.XXX.XXX.XXX:8080/alfresco/cmisatom
dl.store.cmis.system.root.dir=XXXX (this value is setted to company's name)
session.enable.url.with.session.id=true
I would like to ask you if anybody recongnizes some of these symptoms, and if so, what would be your recommendation.
When we access Alfresco from Alfresco app, all goes right. But when we access documents stored in Alfresco through DocumentLibrary things happen really slow. For example, a for loop iterating over whatever 10 documents takes at least 30 or 40 seconds. It's the same if you try with the first or last 10 documents or else, anyway to obtain a single doc the system takes sometimes until ten or more seconds.
Another thing I found is the alfresco access log files: the same request is made along the day until 20.000 times. There are about 800 repeated requests every 24 hours, many of them for the second of third time, but these 20.000 requests, 10000, 9000, ..., are strange for me (Log files are taking about 80 megabytes/24 hour). AFAIK, there are no changes in the Alfresco config besides DB and root.dir values. In addition, adjusting VM memory parameters to change eden, survivor and old space sizes has not had any apparent effect nor behaviour change.
If we use the migration tool in control panel, the system runs for four or five days until VM crashes (OutOfMemoryError, we have adjusted VM memory parameters again without changes).
Processor use of Alfesco machine is many times at 100% (in the production system), but it's the same. It always takes too much time to deliver a document, even with no CPU actvity. No memory problems, no O.S. swapping. Strangely, using Alfesco application all goes as fast as it should.
There are a lot of hooks and portlets, and in some of them the use of DocumentLibrary contains some harcodes and other strange things.
If anybody recognizes some of this, please, any new idea will be appreciated.
Thank you in advance.