Hey - this is from another discussion on saw on here, but I'd like to start a new train of thoughts.
Some users are having much slower page load times when transitioning from Classic to NWE. Tickets have been submitted, yadda yadda, but I'm hoping we can use this thread to maybe come up with some common thing that causes slower loading times for some, but acceptable page load times like what I experience.
But I will need help/input from people with slow load times, and acceptable load times. First, let's start with a list of things we should consider looking at. I'm going to throw a few out there, but please chime in with anything else you all think we should compare. I'm sure there are many of you out there that have a lot more IT/back end experience than I do!
Computer - Mac or PC
Internet provider?
Do you use VPN?
Do you use SSO?
Where do you notice the slowest page load times?
Where do you notice the fastest page load times?
What is your system storage quota, and how much of yours have you used?
What cluster are you on?
Are you NWE, Classic, or a hybrid?
Are you experiencing slow or acceptable load times?
If you are experiencing slow load times, are you experiencing it in NWE, Classic, or the hybrid?
Is there a time of day, work day in particular that you experience slow load times?
What are the time zones your instance works in?
This is a Google Sheet I created, so if you would like to participate, please request permission and I'll share it with you. Let's get some data!
Topics help categorize Community content and increase your ability to discover relevant content.
Views
Replies
Total Likes
Here is some info, hope it helps:
Computer - Mac or PC- PC
Internet provider? Comcast and Spectrum
Do you use VPN? Nope
Do you use SSO? Yes
Where do you notice the slowest page load times? Projects List, Dashboards, Reporting with multiple fields
Where do you notice the fastest page load times? Requests, Custom form completion, Task view
What is your system storage quota, and how much of yours have you used? 456.34 GB of 4,025.41 GB
I'm unsure if this information helps, but our cluster setup is CL05. I have noticed that when there is sometimes and issue that the System Status reports indicate CL05 as the source of most issues.
While we are indeed clients/customers of Workfront, as Admins, it's our responsibility to be active partners in the improvement of this product by providing any and all feedback we can. We can find a solution together. There are too many intelligent and creative individuals in these forums for us not to be able to find viable solutions.
I'm really interested to hear CL05 is the source of the most issues, but is this because we have the most complainers or because it has the most bugs? 😂
I would benefit from answers to additional questions--
I work in both Central Time and Eastern Time Zones in the US. We have teams in just about every time zone in the US and Canada. All experience similar latency issues. As for time of Day, morning, noon and night just about every day. I am permanently in NWE.
Views
Replies
Total Likes
@Mike Haleas‚ @Skye Hansen‚ Hey there - this is a great start. I have added Skye's suggestions to the list in my original post. I have also started a spreadsheet where I can record the answers, and then once we can get a decent number of responses, we might start to see a trend or two. I hope others start responding, too!
Views
Replies
Total Likes
thanks Kelly, I'm not much good at "the cloud", but if it's possible, you can certainly provide a link and we can probably(??) all go there and enter our info if that saves you some time and effort? And this way we can also continue to refine the questions
Yes, that's a great idea. I will put the file on my home drive and share it out.
Views
Replies
Total Likes
Nice thread, Kelly -- thanks for taking the positive initiative!
For those able/interested in sticking their hand further into the technical scorpion lair, I invite you to learn about Chrome's Network Panel, which can provide a wealth of information about all things internettrafficky...and provides a soothing Gantt chart to gaze upon, should the web page you're waiting for take some time to loads.
Regards,
Doug
Views
Replies
Total Likes
I admire the intentions of crowd-sourcing answers to common variables in an effort to identify or pinpoint the root cause of slow performance in NWE amongst ourselves, but am not sure it will bear fruit. I'm especially skeptical that location is a factor.
What I DO recommend, if slow performance impacts your users: perform a series of tests from multiple users working from multiple environments, in a scientific manner. Compile the results and share them with Workfront support. (Not in this thread, unless for some reason you really want to.) This will also record your current performance so you can evaluate if performance is truly better (or not) in the future.
Pick a URL. Everyone that participates in your test should have the same access, so they are loading the approximate same amount of data. For this reason, I chose a dashboard that shows a fair number of records (100-something), and every user that views it gets the same collection of records, so should have similar load times.
Load the URL under multiple conditions:
In Classic, in your computer's standard browser (Edge or Safari).
In Classic, in Chrome and/or Firefox.
In NWE, in your computer's standard browser (Edge or Safari).
In NWE, in Chrome and/or Firefox.
While loading the page in all of the above scenarios, have the browser's Dev Tools open. (This is important.)
In the Dev Tools, view the Network tab. Every browser's dev tools look slightly different, but all should have a Network tab. If you're not sure where to find Dev Tools, Google is your friend.
If you see a long list of URLs in the Network tab, look for a trash can icon or other icon that will clear the list. Clear it, then refresh your page. You will see the list quickly fill up as the page loads, and in the very bottom of the browser, you will see stats on how many requests are being made to Workfront servers, and what is the actual amount of time it takes for the page to load, into fractions of a second. This is MUCH more scientific than a blanket a statement of "NWE is five times slower. (It feels like...)" In some browsers, like Chrome, you need to click a red "stop recording" icon to stop the timer.
Do this a few times in each browser, so you can determine the average load time for the page for each browser in each environment.
When you're done, you should have a minimum of four numbers to show: load time for Browser A in Classic, load time for Browser A in NWE, load time for Browser B in Classic, and load time for Browser B in NWE. If you can test on more browsers, even better.
Along with the URL of the page/dashboard/whatever was tested, provide/compare details about each of the browsers from which each test was performed:
1. OS
2. Browser
3. Browser Version
4. Important: What extensions or add-ons do you have installed on the browser? Each browser has a different method to export a list of the extensions that are installed on it - Google is your friend. This is important because you and your users and me and my users all likely have different extensions installed. Disabling unnecessary extensions can make a big difference in your load times, because there are fewer things processing the data before it's rendered in the browser.
Finally, you'll want to note details about your network connection:
1. On VPN or not?
2. Screen capture of an internet speed test to confirm your bandwidth to the internet is not slowing performance.
The above is annoying to do and takes time, but it is worthwhile. In my case, it gave me a true understanding of the difference in load times and how the environment affects that; as opposed to what it "feels like." Most importantly, I discovered that my preferred browser (Firefox) is consistently slower than Chrome, and Chrome is where I observed the least discrepancy in load times (although NWE was slower in all tests).
Having a collection of real world stats like the above, along with variables like browser and browser extensions, is probably more helpful to Workfront as they work to improve the situation. (I do suspect something will present itself, like, "the slowest pageloads all have XYZ extension enabled - let's explore that")
If nothing else, you'll have true numbers to make recommendations to your users. Like, "Use Chrome, it's the fastest of the slow."
Wow - you have really put a lot of thought into this! I added OS, browser, and installed extensions to the list and put it on the spreadsheet.
For the test you're suggesting, does this mean something like setting up accounts outside of our work instance, uploading dummy files, and then have us all log in and see what our experience is from there?
Views
Replies
Total Likes
Hey gang - here is the link to the Google Sheets file so we can all add in our experiences.
Views
Replies
Total Likes
Views
Like
Replies