save and opening a new work item is take longer time 30 sec
the work item page takes time to load nearly 30 sec.
i have 107 fields in this particular work item and 40 calculated values script attached to this work item
after loading the dependent drop downs take time to display values.
but time taken for a saved work item to open is very minimal leas than 5 sec.
to save and sage the status of a work item it takes 20 sec.
now we have 14k work item created for this project.
is there anything i can look in to for reducing time in opening and saving the work item.
|
Accepted answer
Ralph Schoon (63.4k●3●36●46)
| answered Nov 01 '18, 5:29 a.m.
FORUM ADMINISTRATOR / FORUM MODERATOR / JAZZ DEVELOPER edited Nov 01 '18, 6:14 a.m. This is likely a case of too much customization. Why would one want to have 107 attributes, 40 calculated values?
I tend to call this "customize to death". Typical Symptoms:
I have seen this often enough. I have also seen horrifying questions around 10000 attribute values for one attribute. And, of course, several such attributes which are dependent on each other.
Some customers have thousands of timelines and iterations or thousands of categories. It then takes 15 seconds or more to load all the values to be able to be able to display them to allow changing attributes.
It is quite naive to assume that size does not matter in IT. It does and if all the stuff needs to be loaded, you should not be surprised t takes its time. Slower client laptops are even more impacted as all the stuff needs to be stored somewhere while CPU and RAM is scarce. More users kill off the servers over time.
RTC was designed to support agile development and not to serve as an attribute grave. Although I can understand one could ask for it, for at least some of the requirements, I think there are limits of what can reasonably be expected.
Cases like these are also reasons why there is a lot of push back in development to open up customization capabilities. The assumption is that it is overdone and we get the blame.
In later 6.0.x some effort was put into caching timeline and category data and to cache HTTP filtered value sets in the browser. But there is only so much you can do. All the values need to be initially loaded and that is all using HTTP. Create a HAR file and trace the network so see what contributes. Also check how much RAM is eaten up by the browser.
You can talk to support if there is anything that contributes in addition or if there is anything that could help in terms of enhancements long term. Short term, if there is no other problem the server has, there is only reducing the number of attributed to display and the number of scripts to run.
PS:
If you can automate attributes, why have them? Can you calculate them e.g. for reporting?
kesav d selected this answer as the correct answer
Comments hi hi
i tried
removing few fields and made few attribute hidden
suddenly this error displayed unable to save any work item
com.ibm.team.repository.common.model.impl.ContributorHandleImpl incompatible with java.util.Collection.
pls help asap
How would one be able to help? Test stuff like this in a test system, not in production. And use safe approaches use process sharing and use a new project area to test your changes before changing the sharing.
You can try to rollback the process using the history (Eclipse client, process customization - Process source.
Somehow you try to set a contributor into a list attribute or vice versa.
|
Your answer
Dashboards and work items are no longer publicly available, so some links may be invalid. We now provide similar information through other means. Learn more here.