Why does ArcGIS Pro have to be so slow???

[ericmeyers1]

Why is ArcGIS Pro so slow? To select assets, field calculate, display layers, change symbology... the easiest of tasks that are commonly utilized within ArcMap are a drag on the software.

When will ArcGIS Pro become faster than ArcMap? That will be the day it could replace it as the goto product for GIS professionals.

[EziMolley1]

ArcGIS Pro is SLOW, really SLOW. That's a fact and it means it's not a case by case issue. It's a general issue; every user has noticed and experienced how slow Pro is. At this point, ArcMap remains the number one Esri's products. I believed that Pro was supposed to run next to ArcMap and ArcCatalog on the same machine! At least, that was what we were told. If so, why would 8G works fine for ArcMap and not for Pro? Far from being technical, let me flatly say that ArcGIS Pro thinks too much before executing any type of geoprocessing. It is frustrating. You want to turn a work over to your boss ASAP, ArcMap is the way to go, sadly.  

However, don't get me wrong; I like ArcGIS Pro concept. I use ArcGIS Pro frequently, I like using it. It would have been a game changer had Esri not rushed its release. It is an unfinished software. The base architecture needs some works not patches.  Honestly, ArcGIS Pro is a user-friendly software once you familiarize with it; and it has great potentialities as well. Things you can easily do that are quite daunting in ArcMap: things like editing, 2D to 3D and vice-versa, data sharing, ArcGIS online connection, Tasks, Link Views, and much more.

I would like to finish by saying that what Esri needs to do at this junction is take a step back, re-work the whole ArcGIS Pro project and come up with a more robust and efficient replacement for ArcMap.

[DrewFlater]

Hi Ezi,

Can you describe this further: "let me flatly say that ArcGIS Pro thinks too much before executing any type of geoprocessing"

I am guessing that you mean validation in the tool is taking too long for a tool you are running? After you input some values into a geoprocessing tool, and before the tool is run, the framework evaluates that the values you entered are valid given the requirements of each parameter. If you can provide an example of a tool or tools that take a long time to validate in your workflow we will look into improving them.

Thanks,

Drew

[DuarteCarreira]

Ezi Molley‌ I completely agree with you.

And I'm going to vent a bit more now...

This is not a very specific case scenario where we have to help debugging the situation and provide as much reproduceable data as we can to tech support. arcgis pro does not fit in that situation.

We are not talking about exotic setups or workflows. Everyone knows pro runs badly always. This is basic QC responsibilities and someone is knowingly let the product get out to the public like this. 

Any guy/girl at esri with a common PC will encounter the same issues we are all facing easily.

Please don't tell me you don't know about network performance or you didn't detect the glaring rendering and geoprocessing issues. Don't tell me you don't make performance baseline testing. I know you do. Just look at the excellent System Design Strategies site. It's all there.

So to me this is a question of honesty and transparency. 

So my message to esri is this... own it. 

In truth, arcgis pro is free of charge. So there's a tolerance for a slow, informed, programmed switch from arcmap.

Key is informed...

Stop avoiding the real situation and start showing good leadership.

Tell us what you intend to do about it, even if it's nothing at all.

You could tell us that it's an intrinsic characteristic of the new software, and it will be slower than arcmap for some hardware generations to come, that eventually will catch up. 

Just like going from av3 to arcgis 8 back in 1999 (it was staggering slower).

Tell us to buy real top-end machines, with raid nvme disks, cgi pro video cards, max frequency 2 slots cpus with 20 cores, and min 128gb ram. 

At least we wouldn't be surprised that pro doesn't budge on our measly i7 sata sdd 32gb rigs. We wouldn't waste our money like this anymore.

And/Or you can tell us that you have an optimization plan going on and will include performance improvements in each iteration of the following releases.

And, please, in either case, create a knowledge base document on how to mitigate the existing performance issues. That would be helpful. 

You can start by acknowledging the problem and point to the road ahead to solve it.

Be a leader.

That would win over *all of us* instead of alienating most of us.

[AlexZhuk]

I agree.

Except, the transition from 3.x to 8.x was a breath of fresh air! It was an exiting change!

As I already said in one f these discussions, ESRI would be better off by scrapping the whole Pro thing and starting something new from the beginning (and ditching the ribbon the process). After 4 years of struggles, it should be pretty clear that the Pro is a flop, sorry.

[VladimirStojanovic]

Duarte Carreira, thank you for this post! I agree 100% with you.

ESRI, I am so sorry I cannot get to some quality time to address all issues I have with this piece of ... software you call ArcGIS Pro. Unbelievable how ignorant you are, year after year. I am truly sorry for that.

[KoryKramer]

Hi Vladimir,

I just messaged you.  Please reply to me so that we can collect your thoughtful and constructive feedback.

Thank you,

Kory

[VincentLaunstorfer]

Hi,

This thread is enormous! I do encounter performance issues which are unacceptable. In the meantime, I found an interesting (short) thread which greatly improved speed:

https://community.esri.com/ideas/5157 

I tested it this morning and it is much much better. To me, slowness issues are most of the time linked to GUIs. My issue this morning was about Calculate Field performance. However, I had similar problems editing annotation with Attributes pane open, which considerably slowed things. And I suspect updating symbology is a similar problem but I don't know how to update symobology with no GUI opened...

Hope this help some of you.

[JohnBrockwell]

Get some RAM Bro! 16GB is nothing!

[RTPL_AU]

Hi everyone,

I have commented on the performance of Pro in the past - both on per operation performance as well as the frequent increase in steps compared to Desktop when required to perform a task.
I see Esri is still avoiding the issue and diving into very deep specifics any time an issue is raised. 
My still standing summary to Esri is: Use Desktop & Pro in the real world to perform a range of steps to create maps, data products and related tasks and measure the outcome in metrics such as number of clicks, amount of mouse movement required, GUI draw wait time & process run time as well as overall project run time.
In some aspects Pro is miles ahead of Desktop - I exclusively use it for any workload relating to publishing to Online. For bureau-style cartography and almost all data management I use Catalog & Desktop - often with multiple instances of each open. The Pro approach to editing in the attribute table is deplorable. 

There is a common misconception that you need a serious machine to run spatial workloads in Pro.

You need a serious machine to run Pro. For 90% of my workloads Pro, its overhead and associated serious machine is not required. As an abstract comparison we are getting told we need the latest & greatest machine to run the Steam client. Many games themselves can run on average hardware but without going overboard with ram, cpu & graphics you cannot get to a point to actually play the game..... It may be pretty but if gets in the way it has to go sit in a corner a little while longer.

Kory Kramer  For some people the latest Windows 10 update 1809 may help.

On two of my machines the update has made general network browsing quite a bit faster - diving into a mapped network drive folder in Catalog in both Pro & Desktop is much better after the update. 
Take note that 1809 will break quite a few things in AutoDesk's 2019 products.

[TylerSchwartz2]

Chris Schlebusch wrote:

My still standing summary to Esri is: Use Desktop & Pro in the real world to perform a range of steps to create maps, data products and related tasks and measure the outcome in metrics such as number of clicks, amount of mouse movement required, GUI draw wait time & process run time as well as overall project run time.

In some aspects Pro is miles ahead of Desktop - I exclusively use it for any workload relating to publishing to Online. For bureau-style cartography and almost all data management I use Catalog & Desktop - often with multiple instances of each open. The Pro approach to editing in the attribute table is deplorable. 

This sums it up nicely.  Development team needs to really run both pieces of software side by side and if Pro isnt at least equally as efficient as ArcMap, then they need to fix it.  Running a GP tool to select by attribute/location, apply symbology from a layer file among many other tasks just didnt take that long in ArcMap, nor was it needing to be run as a GP tool... It was near instantaneous and very efficient.

Editing an attribute table is so incredibly clunky and inefficient compared to ArcMap.. ex: it resets the position of the attribute table every time a record is editted, needing to manually scroll back to the next row/column/record which needs to be edited (where in ArcMap hitting return takes you one record down-no need to rescroll).  This is basic software development 101 and should have been caught by now... It makes me wonder what the development team is thinking... 

Its a bummer because Pro has a lot going for it, especially the plethora of new Symbology options available with the additional transparency functionality, and just overall easier symbology editing experience (no more windows within windows within windows).  I also like the new layout and legend UI and find it much better for cartography.  Which is why I still do all my analysis in ArcMap and do map exports in Pro, because ArcMap is faster, but Pro has better cartography options.

Is the development team aware of these issues and just dont know how to fix them? Or are they are just completely unaware of the issues... Im not sure which is worse.  All I know is that Pro should not be less efficient than ArcMap to do the basic tasks I have mentioned - running a GP tool that takes 30 seconds to do something that once took 1 sec is just unacceptable. People are not going to adopt the software if thats the case...  At best, they will only use Pro for certain tasks but still primarily use ArcMap (like me).