First day notes on the general release of SEUM

Some thoughts after my first day on the general release of SEUM.

As background I did run through all the stages of the beta and felt the product was worthy of a purchase.

It appears a lot more of beta dicussion points made it into the general release than I had expected. Some very nice stuff with the avg run times on groupings or locations and cumilative averages.


After day 1 here are my notes on what still needs work.


adding recordings to playback locations is clunky.

hard to figure out where to add them, slow (not possible as far as I've found) to add multiple recordings to a location, or add a recording to multiple sites outside of the creation of the recording.

Even after I figured out how to do this after the fact I kept having issues refinding the method to do it a few hours later. 


x y coordinate fallback

I know we discussed this in the beta and I still think it needs to be on the roadmap. This gives all clients an absolute fallback to recording issues. Its a pain to use but your competitors all use it and in some cases you need to use it.


step wait

Steps within a recording need the ability to wait or pause in cases where you more accurately want to simulate a user. The ability to add pauses after the fact would be a bonus.

This is not an issue of the playbacks proceeding before a step completes but more control factor over how one wishes to run their transactions. Not every transaction is a race.



atlas step level support

This one is critical for me and I started a separate thread on it before. Our current vendor makes visual display of all steps within a transaction very visable and this is something I need to replicate before our final cutover. I will definitely hold off getting more transactions until this is in the product.


issues on install of player

I had one issue installing the updated player on an existing beta player box. did the uninstall, reinstall and got worker process could not launch on every playback. had to reboot, reinstall, etc several times to clear out what ever was causing the issue. It was likely a one off but I thought I'd throw it out there to anyone else replacing beta playback with the final version. It is fixable and the logs were a big help in determining what needed to happen.

The install otherwise went without issue onto a standalone APM box. I did notice when I had the SEUM server running playbacks as well there was a significant impact to http/s template responses in APM. I created another local playback agent and the issue went away.


preventing playback overlap

I am noticing what I believe is overlapping playbacks between my playback locations. This causes session errors when deploying the same recording out to multiple sites. It may be a necessary evil to run each site as a different user to prevent this but it would be nice to see some sort of visibility or control into overlapping transactions to prevent such things from occuring.

I realise as your playback locations increase it would eventually be impossible to schedule them without overlap but it would be nice to have or at least get a warning when overlapping recordings could or would take place.


Overall its a great v1. Savings are impressive against the big boys. 


  • Hello Andrew,
    thank you for your feedback! We appreciate it.

    Now to your points:

    adding recordings to playback locations is clunky.
    What do you mean by "hard to figure out where to add them"? Is workflow confusing for you? As for N recordings to 1 location assignment I understand your point. It's definitely something worth looking on. For internal folks I entered your request under number FB78657.

    x y coordinate fallback  and step wait
    We are considering these features for some future release.

    atlas step level support 
    I replied with workaround to your separate post regarding this.

    preventing playback overlap
    Would "parametrized" recording help here? If you could override recorded credentials used in recording with some different credentials during assignment recording to location so that each location would use separate credentials, would it be acceptable? As you mentioned, synchronizing high number of locations so that they don't play overlapping recordings would introduce much more complexity and could negatively affect overall performance of playbacks.


    Keep giving us your feedback as you want, we really appreciate it. Your comments are always to the point and can make our product better.



  • Hi All,

    Is it possible to insert SEUM jobs to groups on NPM ?

    for example, I created an IIS group on NPM. I expect that the group must contain server and resource usage(IIS is published on this server), IIS APM components also SEUM operation status from other locations.


  • Hi Ramazan,

    yes, this is possible.

    Have a nice day,


  • Jiri,

    parametrized recording as you suggest I think might be the solution to the overlap issue. I do recall from demoing other vendors most had a concept of removing state and user info from a recording. Being able to quickly override x y or z would be most useful.

    on the adding recordings confusion. 

    I have no doubt it was in part due to my workflow. After I reinstalled I created some recordings and then later added in additional agents. So my confusion began outside the normal create a recording add to these locations wizard (which is pretty nice). Looking at ManageMonitors.aspx this morning it seems pretty clear but again if you add in agents after the fact it seemed confusing as how to go about getting them assigned. The N to X again would help out here. Being able either to assign a recording to multiple locations at once (again outside the recording wizard) or assign multiple recordings to a new location at once would be a big help.

  • Jiri,

    I did test your group step workaround and it does work but I don't think it will be supportable on my end. the number of groups simply gets out of control.


    So here is our dashboard that shows apm as well as other data. While most of this data is a summary across multiple servers at the SEUM level what we need to see per instance we host is the 4 step breakdown of our transactions from each location.

    So pardon my poor censoring of the image but player 1 for instance 1 is our 4 step transaction as we wish to view them (using your group workaround) player 1 for instance 2 is what ships with SEUM today which requires another click to drill in to see which step is causing the alert.  Some steps aren't as important as others if they slow down which is why we need to be able to see this break out at a high level.

    dependent on the step that is getting slow then determines which other counters to the right side of the dashboard get important to look at.

    Let me know if that doesn't make sense

  • Hello Andrew,
    first of all - really cool usage of Network Atlas.

    Now I understand that groups workaround is not for you. I thought that you need just a couple of important steps on the map, not all of them.

    Unfortunately current version of SEUM doesn't support this but we already track this as a feature request.

  • great thanks. I will cross my fingers its in for v2. 1.1 would work too :)

    The groups do work pretty nice. I've already got a nice rainbow effect going with my 4 side by side steps. The group math just doesn't add up when each instance is 4 steps x 6 or more locations. and then we x that by 20 or 30 instances.

    I assume there is some limit to group count? My maps already use a ton of groups as I round up in the above example 10 or more servers into a single http response category for instance.

    I can take some snapshots of what our current vendor offers as far as maps of just transactions which include step visibility. I would think it would be a fairly easy dev effort to include that as a new feature since its just a visual hub and spoke diagram that dynamically updates.

    I believe I still have your address someplace. I'll send a cap of what I'm talking about.

  • Sure, send it to me and we'll have a look. As for groups limit, there is no such thing. Only limit is performance of your server that has to process/display groups.

    Again, thank you very much for your feedback.

  • pm me your email? I can't seem to find it in my previous exchanges

  • Very nice.  I am brand new to Solarwinds and SEUM is a part of our initial evaluation.  In a past life I worked somewhere else and used other tools --- including a transaction monitor tool that included a display similar to what you've got above.

    I just wanted to chime in and say that something like this would be a great feature.  I can imagine both operations and application support getting value out of a display like this.

    Seing a breakdown like this at a high level would be great.  You used Network Atlas to generate the screenshot above?  Will have to dig into that piece...