Hello!
We have noticed that the speed of file conversion to PDF has decreased approximately by half. After conducting an analysis, we have observed that the speed worsened around March. Could there have been any updates released that may have led to this issue? Could you please help us this problem?
@AsposeCreatio
Please share some more details of your findings. Please share your input document along with the sample code and expected conversion time.
Hello!
You can find an example code and file for conversion in the attachments. The expected conversion time is 30-35 seconds. However, the current conversion time ranges from 50 to 80-90 seconds.
Code:
image.jpg (44.1 KB)
Files:
Quote_for_test feshchenko_29.docx
Quote_for_test feshchenko_26.docx
@AsposeCreatio
We have opened the following new ticket(s) in our internal issue tracking system and will deliver their fixes according to the terms mentioned in Free Support Policies.
Issue ID(s): WORDSCLOUD-2655
You can obtain Paid Support Services if you need support on a priority basis, along with the direct access to our Paid Support management team.
Hi!
Could you please provide specific deadlines for delivering the solution, as I couldn’t find this information in the aforementioned article?
Thank you!
@AsposeCreatio
As we have recently noticed, the issue is pending investigation in the queue. We will be in a good position to share and update/ETA with you as soon as the issue investigation is completed.
Thanks, I will be waiting for some updates form you.
1 Like
Hello!
Do you have any updates about the file conversion problem?
Thank you!
@AsposeCreatio
Thanks for your inquiry. We have investigated the performance issue at our end. We setup the old (23.10) and new (24.3) versions of the API locally as dockerhub and executed a few requests to convert the attached document several times in each version.
We also performed similar performance measurements on a production server with the current aspose.cloud hosted Aspose.Words Cloud version 24.3.
LOCAL DOCKERHUB SERVER 24.3
4.8 sec
4.4 sec
4.4 sec
4.6 sec
4.4 sec
LOCAL DOCKERHUB SERVER 23.10
4.8 sec
4.7 sec
4.7 sec
4.9 sec
4.6 sec
PRODUCTION SERVER 24.3
30.8 sec
33.3 sec
32.2 sec
28.2 sec
29.3 sec
As you can see, the performance between the old and new versions is no different.
Requests to the production server also showed no degradation in performance and corresponded to the stated 30 seconds from you.
Based on the data presented above, we think your performance problems are caused by the network latency of your internet and are not correlated with our servers.
As another suggestion, please try Aspose.Words Cloud Docker image from dockerhub. It provides you with more control over the cloud service.
Hello!
Thank you very much for your response.
On our end, we have checked the operation of our internal network - no changes that could lead to a deterioration in the speed of file conversion were made. We cannot attribute the issue to the internet connection either, as the problem is reproducible among users in different countries on different continents, whether they are working in the office, remotely from home, or elsewhere.
Please clarify what additional data we need to collect for a deeper analysis?
@AsposeCreatio
Thanks for your feedback. We are looking into it and will share an update with you soon.
Hello!
Do you have any updates about the file conversion problem?
Thank you!
The issue is still under investigation. We reproduced the performance downgrade for the first request, and then it performed as we tested before for further requests. Once we get more information, we share it with you.
Thanks for your request.
Hello,
Could you please provide an update on the progress of analyzing our issue?
I am afraid we couldn’t reproduce the exact issue you faced, but we are doing our best to replicate the issue.
Could you please specify if we can provide you with additional information to help analyze the issue? If so, please provide a list of the required information.
We did load testing. The sequential chain of requests shows expected results in a range of 30 to 40 seconds per request.
Success in 36,9053552 sec.
Success in 39,9640732 sec.
Success in 33,4823562 sec.
Success in 36,3312228 sec.
But if the requests are executed in parallel, the execution time for one of four parallel requests increases to 90-100 seconds.
Success in 42,4944265 sec.
Success in 38,1358389 sec.
Success in 66,6146987 sec.
Success in 97,137908 sec.
So, the reason for performance issues is that your requests execute during high load. Could you please try running requests at different times to check if this is the case and share the results with us?
Also, please share your CusomerID so we can track your requests in logs.
As another suggestion, please try Aspose.Words Cloud Docker image from dockerhub.
Hello,
Apologies for the delayed response.
I have conducted testing on file generation in our production environment at various times and on different days. Regardless of the time of day, it appears that our file conversion time remains consistent, ranging from 40 to 45 seconds, and occasionally up to 55 seconds.
Here are the measurement results (generation time indicated in GMT+3):
- 22.04.2024 17:32 - 47.7 sec
- 23.04.2024 9:19 - 49.1 sec
- 23.04.2024 15:30 - 49.7 sec
- 23.04.2024 16:32 - 55.2 sec
- 23.04.2024 17:40 - 45.2 sec
- 24.04.2024 8:53 - 45.1 sec
- 24.04.2024 10:44 - 40.9 sec
- 24.04.2024 12:16 - 49.2 sec
- 24.04.2024 15:16 - 39.7 sec
The closest we’ve come to achieving a generation time near 30 seconds was on a weekend - 28.04.2024 22:42 - 33.1 sec. However, I was one of the few users operating on the environment at that time.
Our Client ID is 783e8c50-5e70-401e-b571-6406d864ec30.
We kindly request your recommendations on how we can improve the file conversion speed, considering that multiple users are utilizing our environment simultaneously.
Thank you,
Juliia
@AsposeCreatio
Thank you for your feedback. Could you provide a sample minimal project to replicate this performance issue at our end? Additionally, have you considered the self-hosting solution using Docker? It can significantly improve the service performance.
Hello!
Apologies for the delayed response.
Could you clarify what do you mean by “sample minimal project”? Do you need an example file?
Thanks!