nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Diff and Merge Performance Issues

medini_1028038066

Status : feature requested with Project Owner Priority : Major, #1 Schedule : (according to Project Owner) Current state : very low/no performance even with high RAM usage Critisism on current state : not usable in daily work Objective of the change : diff and merge is needed during the daily work Proposed change : increase the tool performance Original Author,Eva Baur: "the performance of diff and merge is way too high, more RAM is already used. Question: to reduce the performance is there a possibility of only chapterwise or a specific diff& merge possible, line #94"

Find more posts tagged with

medini

medini legacy

Accepted answers

All comments

medini_1036016653

In the meanwhile I further explored how/if I am able to influence medini's ram usage. Apart from setting the maximum allowed ram usage in "mediniAnalyze.ini" via the flag -Xmx10G there is actually a way to manually invoce garbage collection (GC): One has to enable the preference General -> "Show heap status": wow! :-) From early experience it helps tremendously to click the garbage tray icon once after a "Compare/Merge" operation is done or a project is closed. Please (!) consider this an automatic task, once a user returns to the "Analysis" view or closes a project. Thank you! Peter

medini_1036016653

(There was obviously an issue pasting the screenshot in my previous post).

medini_1036016653

Sorry to post yet again... What is interesting though is that at least my task manager is not aware of any freed ram. It currently sits at 5GB! I observed the usage climbing by about 1 GB each, after each "merge subtree changes". Best regards Peter

medini_1036016653

Ich hatte lokal noch eine Antwort (mehrere Monate alt), die ich damals wegen Problemen mit der Forum-SW nicht abschicken konnte. Hier noch ein Versuch: -- Sorry for bothering again. Even with v2020R1 I seem to
be unable to delete the tree in an efficient manner. It seems the medini
process isn't even working (hard) on the task: 8% CPU utilization In case you wonder about my laptop's hardware: https://ark.intel.com/content/www/de/de/ark/products/134899/intel-core-i7-8850h-processor-9m-cache-up-to-4-30-ghz.html -> 6 Cores / 12 Threads 100% / 12 Threads = 8,33 %/Thread -> hm It seems that medini is mostly making use of single
threaded calculation/operations. Now I'm having the practical issue of deleting a tree
:-) Admittedly it got created using a script, but it actually only about 1/30th
of the tree I intend on generating Monday. Would you please be so kind and get
in touch with me? I try to supply yet another stripped-down example
project for you to test and will attach it here. With the weekend approaching I took a last resort measure, googled for heap
settings that could be applied and arrived at -XX:+PrintGCDetails -XX:+AggressiveHeap The aggressive heap (https://docs.oracle.com/javase/8/docs/technotes/tools/windows/java.html) surprisingly
did the trick! I tested both v2020R1 and v2019R1.1. Oracle states: "
Enables Java heap optimization. This sets various parameters to be optimal for
long-running jobs with intensive memory allocation, based on the configuration
of the computer (RAM and CPU). By default, the option is disabled and the heap
is not optimized. " I'm no expert on Java. Could you please clarify, if further options might
make sense to set, e.g.: -XX:+AggressiveOpts
-XX:+UseLargePages
(only Windows Server 2003?!)

-XX:+UseParallelGC -XX:+UseParNewGC

medini_1036016653

Zur Dokumentation hier noch zwei weitere Links, die sich mit der "Optimierung" der Java Garbage Collection (GC) befassen: https://www.jvmhost.com/articles/what-is-java-lang-outofmemoryerror-gc-overhead-limit-exceeded/ https://www.oracle.com/java/technologies/tuning-garbage-collection-v50-java-virtual-machine.html Könnte dazu bitte jemand Stellung nehmen?

medini_1036016653

This report in my opinion is still of very high priority. Diff&Merge is a continious "pain point" during work with medini analyze. While with v2020R2 the process got somewhat streamlined (warning message can be permanently disabled), now another dialog pops up at least once ("warning about scripts", possibly due to FTA events with JavaScript). We often experience stalls/crashes during diff&merge operations. Reporting those issues and following up is time consuming. General improvement of speed/robustness would be highly appreciated. Best regards Peter

medini_1028038066

Dear Peter, the development team confirmed, that the feature request we have raised based on the discussions in this thread are continuously considered and we will keep you informed, in case major changes are available, thank you for your patience.