Laserfiche Tutorial: Is Your Laserfiche Workflow Overheating?

[Music] Hello and welcome everyone to this CDI webinar titled is your laserfish workflow overheating. My name is Kyle Neble technical trainer here at CDI and I'm looking forward to presenting some interesting and uh neat topics regarding laserfish workflow. Uh the question is, is your workflow overheating? Uh what we'll do is we'll take a look at a four different ways to check on workflow and make sure that it's running optimally. Uh first off, we're going to search workflow from or for problem indicators. Uh we'll also look how to prevent a runaway workflow, show you how to build that into all your workflows. We'll take a look at some administrative tasks like how to quote truncate logs in laserfish workflow and then we'll also look at some uh SQL server or SQL server optimization for uh workflow in regards to uh database indexing and rebuilding them. Uh little housekeeping items for our session here today in the panel. uh you'll be muted but there is a question box and the panel there uh will allow you to post questions in there uh for that purpose. Um if you want we can provide screenshots uh but as you can see it's a fairly simple presentation so no big deal. Uh and the really good news is we are recording the session and we will get this up on our YouTube channel. Uh and therefore we will be uh posting that and you can watch this later on or send it to your friends who might be interested in this topic as well. Uh who's our target audience? Uh well definitely laserfish admins. Uh those of you who support your laser system and build workflows, you'll find this really interesting. Um, any IT staff who tangentially knows LaserFish might be interested as well or if they support SQL Server might be interested in seeing some things that we do with workflow. And if you're just a Laserfish subject matter expert and just want to learn a bit more, this will be a little bit of a under the hood look at some of the things in Laserfish that can help us make sure it's running optimally. So, uh, we're going to look at some preventative maintenance and optimization tricks here today. Uh, when it comes to problem indicators, uh, normally you your workflows run just fine. The uh workflows may have certain errors or warnings or even uh terminate which uh at some point within that workflow process some activity was not able to complete successfully and potentially has stopped the workflow right there at that uh moment. So it doesn't continue and does not complete the workflow. So a terminated workflow uh could end at the last activity or somewhere in the middle of all the activities and cause that workflow to stop running and end prematurely and we don't want to see that. However, there are ways to find uh those workflows and we'll take a look at that here briefly. Uh the example screenshot here we can see that uh you can search in the workflow designer tool for a workflow status, any errors, warnings or messages. uh we can search for all workflows or a specific workflow that has any of these properties. It's a really great tool. It's almost like searching in the repository for content, but yet it's related to workflow uh in the workflow designer. So, a really great tool for searching for problems that you might be uh finding in workflow. Another topic is how do we prevent a runaway workflow? Uh lots of workflows are going to start automatically. Uh, some you can start on your own manually if they're been designed that way, but most workflows will trigger based on a set of rules, but we don't want it to run unless some actual user has possibly added a document or changed it or done something in the repository. But we don't want workflow to trigger itself. And we'll take a look at how we add this simple condition that will stop runaway workflows. Uh and we just tell it that the user should not equal workflow or the WF account or the WF agent whatever you've used as your workflow user that is taking action for us. Uh we can tell it not to use uh that as a trigger for starting workflows. Now uh I will say that uh most of the content here today is specific to onremise workflow. uh laserfish cloud workflows have their own method of stopping workflows from running away. Uh it's a different environment completely from an on-remise system, but it's also got a built-in uh way to limit runaway workflows as well. So we don't have to worry about that too much in a cloud environment. So all of the topics and concepts here today are really specifically for on premise systems. uh a runway workflow could also potentially be uh found and the scenario is in fact I had a client many many many years ago who hadn't seen my training or had learned about this had a system where uh the workflow was calling uh an action that created another entry that then ran back and triggered a second workflow that looped over and over between these to workflows at which point it had created uh quite a few log entries. In fact, the uh workflow logs the u entries that it was creating uh basically filled up the server and caused the server to crash because it ran out of hard drive space. So this looping it took about a week to get into that problem. But with that potent that way it was configured uh the two workflows were combining together to loop so fast and so long that within about a week it had locked up the server because it had run out of server space. Now unfortunately in that particular scenario their system had been set up on the C drive of the server which really was bad because then the operating system couldn't run. Uh we did recover uh and get the server back up thankfully, but it took a little bit of effort and it's something we just don't want to see you uh run into. So we'll show you how to set this up in the workflow designer for your workflows. Uh I mentioned how you can truncate logs in laserfish workflow. Workflow and its engine has log files that we can use to run reports uh look at instances that have run in the past and basically if we want to troubleshoot historically speaking but sometimes you've got so many workflows that are running so often that there's really no point in keeping them longer than you need to. Many of our clients keep uh their logs only for about a month or two uh especially with high volume workflow scenarios. So we'll go ahead and take a look at how we can change that up. Here's an example of uh within the workflow administration console where we can set particular settings and I'll show you how we get in there uh in a moment. All right. And then finally uh not necessarily in workflow but the workflow database upon which workflow relies. We can optimize SQL server by taking action in the indexes whether we want to either reorganize or rebuild the indexes in the SQL database itself to make that system run fast. Now this is I should say really only for those high volume workflow environments. uh if you're you know running 10 workflows and they run a hundred times a day, it's going to take quite a few months to years to get to the point where you actually need to even consider a reorg or a rebuild of your indexes. High volume systems certainly uh we might want to consider this as an option. So we'll show you what that looks like in the tool uh that we might want to run. Now, this is also probably more uh something that an IT administrator would be doing for you rather than uh if you're just a workflow or laser u subject matter expert or workflow designer, you might not want to touch SQL Server itself. But uh we'll show you what that looks like as well. All right, so that being said, let's go ahead and take a look at a few of those uh items in there. And the first one is the uh search for workflows that might have trouble. And I'm in the laser 12 workflow uh designer tool right here. And uh within the search area and if it's not available, you can go ahead and do that. You have a search tab. If this search uh tab is not visible, for example, you can see this down at the bottom left corner. I've got a toolbox, a search tab, and a common searches. If they're not visible, go ahead and click on the view option and toggle it on. Uh and you can add common searches or the uh advanced search within workflow. Uh and I've got both of those on. So, let's take a look at common searches first. And in fact, up at the very top in the common searches, you could look for today's terminated workflows. Hopefully, you don't have any, but if you wanted to find any, that would be a great way to do that. But that's only going to report, as you can tell, it's a pre-built report there. You don't have to build them yourself. So, you could look in there and get some results. Now, I do have some results. I just ran a test search and I do see that I've got a couple workflows that have uh some messages. If I see a red circle with an X in it, that tells me that that's a termination event. And we'll have to figure out what happened in there uh and troubleshoot that. So, that's a result of a search that we see here in the search results tab. if that tab isn't there but then you run the search it will then appear and show you the results and it'll stay there from then on. So that's in the common searches but in fact uh this search results was found by using the advanced search or just the what they call search box and as you can see in the top right from the top we have the workflow status or state and you can mix and match whatever you want using these filters. uh in this case I searched for terminated and then uh any errors, warnings or messages. I didn't actually select any so those might not happen. So you could toggle them on if they're not toggled by the way they actually it just kind of expands that area when you toggle it and then you have to select one of the subheadings under each of those items. Uh so in this case I had checked terminated uh errors, warnings and messages. I actually have not selected anything. Uh the time of last last action is really probably a good choice to make here as well. And then if you want to be specific about your workflow names, you can choose that. And you can even choose what version of that workflow it is. Uh you might be editing workflows and you may edit it once a month. of or not that you edit workflows on a regular basis, but maybe the latest version. Uh you might want to choose that or it doesn't matter which version you're looking for. You could choose those as options. And the filters just keep going on and on and on. You could have a workflow start date or a rule that uh also was part of that workflow. Your workflows might have several different start rules, right? So again, lots of different ways to search. And in this case, uh, having set this up, just hitting the search, that runs the search and then shows us our results in the results screen. And then, uh, in this case, there's two instances showing up in this list. Uh, an instance is just a single workflow that ran, uh, against maybe a single entry uh, or it's just the workflow that ran at that particular time. And if you open it up by double clicking, the workflow designer tool, which you probably are familiar with, will show us where the termination occurred. And we can see following the path, it dropped into an activity that at this point terminated the workflow and never continued past that point. And then you can spend the time figuring out why did it not happen. Uh, normally you'll probably at least initially look at the messages tab there at the top and figure out why it could not insert. um in this case right whatever the particular activity might have been. So that is how we can search for problem indicators and then drill down into the workflow and troubleshoot from there. Uh once we fixed all that theoretically our workflows are all good. All right. Now the workflows themselves probably have start rules and the way you find those are in the rule manager. So if you have rules uh enabled on a workflow and in this example if you're looking right here we can see we have the uh autofile workflow for finance and it has two start rules next to it. we can see if a document has been created or entry been created or uh the other event is entry moved. So these are triggers that start the workflow. Uh and we built some rules, right, that uh tell us what's going on. And if we edit one of the rules, doesn't matter which one, but they all should have the same thing that we're going to look for, which is our uh specific item that keeps us from having runaway workflows. So, regardless of what all the conditions are that make this workflow start, these are the start conditions uh in this editor. The one item you want to be putting in there, I think it's pretty important, unless you don't care, but I tend to do this all the time and I recommend you do as well, is add the uh user does not equal and then the uh account used to run workflows. Normally, this is not a licensed user. It's an unlicensed account. You can create it uh with any name you want. Doesn't need a license, but you're probably going to need to give it access to your repository. uh and all the folders in which this workflow needs to run. However, um this tells the workflow uh start rule to ignore any events that were done uh by the workflow account itself. So theoretically this workflow or or other workflows could trigger this workflow again um and therefore it would ignore those because only end users theoretically would be making changes uh or creation or whatever the event is in the repository. So that's going to be a great tool here for on premise environments where you set that up and it'll ignore any situation where workflow is trying to do something and then potentially start another workflow uh which then could get you into a loop. Really great tool. Highly recommended. I do this all the time. And even if you don't think it's useful, I would suggest adding that in there. Um, it's a if you don't have it in there and you want to add it, just hit the add condition button and in the dropdown there's a user choice right there in the drop-own list. Go ahead and usually it says um uh does not equal and then whatever that account is. I've seen all kinds of names in here. uh WF agent, WF for short, whatever you call it. That account is probably found in the admin console uh for laserfish. Uh and basically it's a service account that really nobody logs in with, but it is going to behave as a user in the repository to do things, but it's also something we don't license. And then therefore, we can tell it uh tell the workflow start rules to ignore anything done by workflow. All right, great stuff. And then publish that, uh, addition and you're good to go. So, uh, and do that for all start rules. They don't autopop populate in each one. If you change it in one, it's not going to add it to the other. You have to go into every start rule and make sure that that is in there if that's the case, and you do need that. Okay. All right. Uh the next item is the way to truncate logs in workflow. Now this is actually done in a different area not in the workflow designer but if you go to tools the workflow administration console is there in the drop-down list and this does normally get installed when you install the workflow designer on your workstation. Uh, so you can then open that up and go view that tool. All right. And I'm going to cancel out. That launches the tool right here that we see our admin console. And even if you have version 12, this all looks the same. So, uh, you'll see the server name uh, at the root there. You'll see all the different sections in there. And what we want to do is go down to server configuration. Right? So you'll expand the server configuration area, highlight advanced server options, and you can just click properties on the top right or doubleclick on any of the items in here. And when you do that, it pops a dialogue box. And the first thing it comes up with is the maintenance tasks. Um, and the second tab you might want to look at is log files, but typically the maintenance tab is where we want to go. And if you drop this down to let's say 30 days, um, for removing completed instances, um, unreferenced workflow, message cache, all those, uh, you can, if you say 30 days for all three of these, well, the timeout would be, well, the last one, don't worry about it. But these first two, if you only want to keep things for 2 months, make it 60 days, 65 days, whatever you want. Uh, and that will then keep the log files from growing too large. Uh, and again, this really, I think, would be useful in very high volume workflow environments where you have uh quite a few workflows running all the time, all day long. Uh so again for a quiet system maybe this is fine but I would say three months uh for a busy system might be just fine. Uh only because if you are having issues you will find them faster than if you have a quiet system uh and that issue only crops up every six months. Uh so you probably want to keep your logs for longer in a quieter system than in a more active system. It's kind of a interesting relationship there. All right. So, that's kind of automatic maintenance that workflow uh will do to keep uh your log files building up uh and causing any undo storage issues. Granted, uh these days with storage being so cheap and with expandable storage systems, not that big of a deal. Uh but again, one way to keep your logs in good shape and not growing too outsized. The log files tab, uh you can also set it to compress log files or even delete uh old log files at a certain point uh however uh you want to do that as well. Okay, so those two items are pretty handy and very easy to set. There is a help button to describe more about the minutiae about all these. I would say that's a really great place to just keep things nice and uh trimmed out and not overgrowing your logs uh for workflow. Okay. Click okay when you're done setting that and that will be maintained. Do you have to restart the workflow server service? Uh I don't believe so. uh couldn't hurt but um normally those settings are implemented as you set them. Okay. All right. And then the final topic is something outside of workflow but relative to workflow itself. There is normally when you have workflow installed on the database server the Microsoft SQL database server there's usually a workflow database and I'm showing one in my instance right here. You can see that listed right there. And there are what they call indexes that help the database get information back to the querying application or the workflow engine uh rather quickly. Uh it's kind of like um a table of contents for information and it says hey you know if it's if you're looking for this go find it here kind of thing. It's a little helper that makes your system run faster when doing uh queries uh in the background. Now the application themselves like the workflow application it is doing uh queries to the workflow database to get information that display in the workflow designer or in things that happen in your database uh and in your repository so that um things just work well. In fact, what we're going to learn here today on reindexing, you could reindex your laser fee uh servers indexes as well and uh do that. So uh each database table so within a database there are tables that contain information and within any one of those tables uh let's see let's look for one. Here we go. We'll just choose the workflow table within the workflow database. Go figure. And you'll notice there's an indexes folder. And there's a couple indexes in here that help the system do their thing. Uh now, before you do any of these tasks, and I would suggest if you're not comfortable or don't understand what this does, read up on Microsoft's uh SQL database uh site and learn more about what we're doing here. uh because you wouldn't normally do this on a single table like I'm showing here, but I'm showing you that you do have options here if you think that one of these indexes is causing you trouble. So, it's a kind of a nond. It's not super uh difficult to do. uh but I would recommend doing the following before you even do a reindex or a rebuild of those indexes and that is to do a where's my tasks backup. Backing up a database is going to back up the indexes as they are right now. So if you realize that oops something went really wrong, you can restore the entire database back to the state it was prior to you doing this particular reorg or rebuild of the indexes. Okay, so I cannot stress this enough. Backups of your SQL databases are probably the most important thing u aside from the file system uh itself of a laserfish system. So if you are going to do this, make sure you have a very recent backup that you can revert to if things go sideways. Typically it won't, but you never know. Murphy's law sometimes can come up and strike when never expected. So the uh backup of a particular database could be done. If you have a nightly backup, great. uh make sure that's done and that you can tell that there is a recent backup available. Having done that, you can then either run scripts or I have this I'm using what they call the Microsoft SQL Server Management Studio tool to look at some of this information and show you what's going on. Now typically you could uh run a script as they call it that could look for all indexes in all databases and do a reorg. Now a reorg will just basically make sure that all the pointers are not out of order u within the index itself. Again that index is just saying hey if you're looking for this it's going to be found in this view or this table or um this column. So, it just basically is a kind of a table of contents, if you will, just to keep it simple. But they can get out of order uh as time goes on. And so, a reorg is the first thing you might try before you actually do a complete rebuild. These indexes, I'm not sure if I can show exactly how big it is, but uh under properties, uh we might be able to take a look and see what it has. There it is. Uh and then the uh info uh we can see the size uh right there and not sure what that indicates. Not necessarily megabytes, but there are uh numbers that might be larger than that or smaller than that. Uh it just depends. But if you want to individually do this, uh this tool does give you that capability. Uh so you've got your choice of either rebuild or rig or organize on that individual item. All right. Uh now on the workflow let's see we do have the workflow itself and the workflow index is there. Uh if you don't select an individual index you get the rebuild all or the reorganize all. And again choose reorganize first uh before you do a rebuild. Uh I would say reorganize is less invasive. You can do this while the system is running. Uh but I would suggest if you're doing any of this and you have a busy workflow system, do this in off hours. Uh it shouldn't take that long, but we don't want to interrupt uh any active workflows. So if you have a quiet time in the evening, I would suggest doing it then. All right. So one of those choices is usually good. uh don't disable anything even though they kind of show it there. No, don't do that. Uh those indexes are important. Not all databases have indexes, but um you could also run a script that runs against all that. Uh if you want to learn more, Laserfish is sorry, Microsoft's SQL Server site has quite a bit of information or just search on the internet. All right. So, those are the four ways that we can keep our uh laser fish workflow system running optimally. And at this point, I open it up to questions for anybody. I haven't seen anything come through at the moment, but hopefully uh that does stir your IT or creative juices as far as maintaining and optimizing your laser workflow. All right. Uh some helpful resources, our CDI website is uh CDI.support. Uh we do have a YouTube channel. Uh, as I mentioned, this webinar will show up on our YouTube channel uh, in just a day or two. Uh, we do webinars all the time. Go ahead and go to our support our homepage cdi.support and look for the webinars link and you'll find uh, a way to sign up there for any future webinars. We love having you and we do this quite often. Uh we generate uh tips on our blog and we have those going back uh many many months and many years uh with little tips and tricks about using different laser fish products. We put out a newsletter as well every month and we look forward to seeing you that there. Uh and if you don't receive it uh go ahead and sign up uh to receive that. Laserfish has quite a few options uh available. All right. So the um uh laser fish info here is uh they do have training uh for certifications that um the training is free. The certification courses are a paid for item. So that's pretty cool. Um they do have also some getting started guides which are really cool. Um they actually solve some scenarios. So they'll actually use workflow and some other tools to uh solve a business uh problem that are described in these um getting started guides. So a really handy tool for a lot of us who have just learned or just become uh exposed to laserfish. So if you're learning laser fish, these are really great tools. Also really great for learning Laserfish is the answers forum at answers.laserfish.com where you can see questions that have been posed uh by numerous uh folks about how to do something within laserfish and they have it broken down by every single product and you'll find a just plethora of information in there. I've solved my own problems and I know a lot but I've been able to solve problems just by going to answers. com as well. Uh this information is usually accessible by a um free support site account. So sign up for your free support site account uh because some of the access is not there. All right, I did see a question come in and uh the question is for workflow starting rules uh is the Windows task scheduler sometimes running workflows even though the workflow designer shows that the rule is disabled? That's a good question. I am not sure that it would be doing that. The workflow task uh sorry the Windows task scheduler would normally be for quick fields u storage of documents after hours even though the quickfield session has processed documents it calls a work a Windows task scheduler to store documents which might be triggering a workflow after that because the workflow triggers don't have a time frame it just knows content came into the repository And if it's set to start a workflow when entries are created and the Windows task scheduler had stored quickfields process documents into the system, that could start the workflow. So that might be what what you're seeing. But if that's not the case, uh we could dive into that. Just open a case with us um and we can take a quick look. Uh, and then there we go. Uh, question on the question as to where the workflow log files live. U, I'll get that back uh, an answer to you there. I don't know off the top of my head. So, I'll get that to you, Danielle. Uh, no worries there. And then if you're using import agent um and the import agent files go to the import agent error folder um import agent usually just tries to put documents into a repository. And if it's failing to do so, it's either u because the repository wasn't available when it tried to do that uh or the um maybe the metadata that it was trying to assign was out of range. Maybe it was set to put a value and these are all static values onto a document and the value is larger than the field supported. So maybe somebody changed the field size to support less than the original width of characters and you've got a default in your import agent that's more than that. It's a it's going to cause that problem to occur. So troubleshoot import agent uh on that and it'll it'll the import agent errors are found in the event viewer on the Windows server on or the Windows machine on which that import agent is installed. So we'll do that. All right. Um so uh that is some great questions. Really appreciate that. Again my name is Kyle Neble, technical trainer here at CDI. If you have any questions, uh, shoot me an email. If you want to set up, uh, consulting meetings, our professional services.sup support, uh, email is good for that. And if you have SA uh, sales questions, sales@cdi.sup support. And if you need to open a support case for errors that you're seeing, uh, just reach out through support cdi.support support or on our main uh website we do have a phone number uh where you could potentially reach out as well. All right, we thank you very much for your attendance today for our webinar and we look forward to seeing you at the next one. That concludes our session for today. [Music] [Music]