Hello,
as stated in the user forum I am new to openkm and reviving my j2ee skills from tens of years ago.
I was now uploading about 1700 files to a single folder which is meant to be the "Unhandled" folder for REST uploads.
When I try to view the documents it takes a long time until they appear. A lot of the time might get lost due to DEBUG mode and file logging.
But when I digged a bit deeper I realized that even when I have enabled pagination and only show 10 entries per folder, thousands of SQLs are executed for the short period of time to show the first page of this folder and then immediately go to second page.
In that (lets say 1 minute) there are 21.970 result set processing entries in the hibernate trace file.
For example:
Maybe my approach is wrong. But probably there is a limitation missing if pagination is enabled. Can you please take a look?
For the attached file I started tomcat in DEBUG. Then did wait until it finished to start up. Pagination in openkm is already switched on.
1. I clicked on "_New" folder with approx. 1700 documents.
2. I clicked on little right arrow to go to next page
3. Stopped tomcat right after displaying the results.
After that I cut off all the lines until the first statement to get the node_document entries.
Thanks a lot.
EDIT: The forum says: failed to move uploaded file. So I cannot upload it. I think the description to reproduce it is sufficient. If not let me know and I can send via email or other option.
as stated in the user forum I am new to openkm and reviving my j2ee skills from tens of years ago.
I was now uploading about 1700 files to a single folder which is meant to be the "Unhandled" folder for REST uploads.
When I try to view the documents it takes a long time until they appear. A lot of the time might get lost due to DEBUG mode and file logging.
But when I digged a bit deeper I realized that even when I have enabled pagination and only show 10 entries per folder, thousands of SQLs are executed for the short period of time to show the first page of this folder and then immediately go to second page.
In that (lets say 1 minute) there are 21.970 result set processing entries in the hibernate trace file.
For example:
Code: Select all
This is there once per document in folder. Not only for document to show in pagination. Same for properties, subscriptions etc./* load collection com.openkm.dao.bean.NodeBase.rolePermissions */
select rolepermis0_.NRP_NODE as NRP1_83_0_, rolepermis0_.NRP_PERMISSION as NRP2_0_, rolepermis0_.NRP_ROLE as NRP3_0_ from OKM_NODE_ROLE_PERMISSION rolepermis0_ where rolepermis0_.NRP_NODE=?
Maybe my approach is wrong. But probably there is a limitation missing if pagination is enabled. Can you please take a look?
For the attached file I started tomcat in DEBUG. Then did wait until it finished to start up. Pagination in openkm is already switched on.
1. I clicked on "_New" folder with approx. 1700 documents.
2. I clicked on little right arrow to go to next page
3. Stopped tomcat right after displaying the results.
After that I cut off all the lines until the first statement to get the node_document entries.
Thanks a lot.
EDIT: The forum says: failed to move uploaded file. So I cannot upload it. I think the description to reproduce it is sufficient. If not let me know and I can send via email or other option.