STT - Migrate Qlik Sense Enterprise Like a Boss

[Music] hi everyone my name is Mario petre I'm a senior technical support engineer with click support I uh work out of the Lund office and been with a company for the last five plus years I hope to uh share with you some tips and tricks on migrating your click Send site to a new environment today all right the agenda for today will be um reviewing the backup and restore procedures as well as uh thinking about some planning transferring files ahead how to manage your certificates how to transfer them correctly from sight to site uh depending on which server you are planning to migrate uh we'll discuss any pitfalls during this process and uh some tips and tricks and some steps for automating now Mario my understanding is basically when everything goes right it's just back up and restore that's pretty much how things should go and everything smooth where can we find all those specific steps though well you're absolutely right Troy um we have a step-by-step detailed guide on backing up and restoring a site as well as transferring the content over to a new site uh on our uh online help page okay so I see we're here on help. click.com under click sense for administrators and we've got the most recent version September 2020 yeah this page contains step-by-step uh procedures on how to back up your certificates what other considerations in terms of data transfer you need to pay attention to such as moving your app content and all the procedures necessary to restoring that site these instru can be used either for the purpose of migration like we'll be doing today or as a periodic backup procedure yes backups are are wonderful because things don't always work as they're supposed to real quick what is the scope of what we're talking about today we'll be walking you through the process of migrating the central node of a fairly simple environment we have a central node we have a service cluster and we also have a rim node added to the environment we'll be doing this in Virtual machines in this particular process it's a VM to VM transfer the same process would apply if you're moving for example from virtual to physical or vice versa or actually uh migrating into a cloud provider and why would somebody want to migrate in the first place there's a few things that can drive that decision environment scaling I I would say is the primary one okay why don't we take a look at the environment we're going to be migrating from what I have here shared with you is the central node of the current environment let's go ahead and take a quick look at the qmc and what version of Click senses this this is our latest version click September 2020 we are running on Windows Server 2016 on all machines um as you can see this machine has gone through a number of migrations and different versions of the monitoring apps uh but we also have some sample uh applications here there is a data connection currently to a postr SQL Server we also have a rim node available a few Custom Security rules to give access to a stream and this is how the Hub currently looks like so we'll be shooting for like for like transfer of content and metadata on a new server called sent server 2 while the old Central node will be quote unquote decommissioned it will no longer be available okay Mario so how should one get started before you actually run through the process there's quite a bit of planning that I like to do before tackling one of these operations first of all I want to know exactly what's in the current environment uh and where things are setup and the qmc can tell you pretty much everything you need to know about that Windows itself can tell you everything you need to know about the accounts and uh the security for these we'll get started on the online help that's where I start as well uh especially for the backing up and then restoring the database we don't expect you to remember these by heart I highly recommend that you familiarize yourself with the steps involved and make sure that uh you have a simple checklist in place for both the original servers as well as the destination just to to make sure that you run through these steps in a in an organized fashion so I have some questions Mario we're going to be migrating and correct me if I'm wrong all of our apps all of our tasks all of our rules that are set up in the qmc will the domain be the same between the two environments in this particular instance yes in the case of the same domain the process ends up being a lot simpler your users will remain intact however if the domain changes there are some considerations for example if your user IDs will remain intact on the new domain let's say you have the exact same user John do exists with the exact same john. do user ID in the new domain but only the domain part changes uh there are ways that we can approach this uh you can uh contact Professional Services for example they can help you with a query for your backend system to uh update all of these records in bulk and resync the accounts to the uh to the new domain if they already exist in active directory it should be as simple as rerunning your user directory sync task once you have modif ified it with a new domain controller address so that it remaps these uh users in the background is that sync task automatic the user directory syn task should exist if you are switching domains my recommendation will be to set up a brand new user directory sync task and configure it with the new domain will the account running the services change that will remain the same however if you need to change it we need to make sure that the new service account has full control file permissions on the service cluster the account uh has permission to log in as a service on the new server under local policy and user rights management and the same account also has to be a member of the clicksend service users group on the new server optionally you also need to add it to your local administrators group but of course uh you can run the services without local admin rights if you need to upgrade your click sense server software say from version 3.2 to a more modern version like September 2020 as well as upgrading the environment which should you do first the software or the hardware I would tackle the software first um mainly for one reason you you have probably run click Sense on this environment for quite a while um this environment is well known to internal IT staff its configuration is well known and hopefully very well documented and if something should go wrong with either the upgrade or de migration it would be easier for for support to troubleshoot on an existing system that has no variables and known configuration than troubleshooting on a brand new system that makes sense okay so as a first step as a reference in the guide the first thing we need to take a look at and make sure that we have a a good backup of our deally sent certificates in this particular example we're going to be also U moving the central node to a new host with a new name that requires us to issue a set of certificates from the qmc as well as backing up the current central node certificates manually uh an example of this type of data that is encrypted using your certificates would be connection strings one example of a connection string that would break without the proper transfer certificates would be this this database is running on a separate machine when the engine tries to access those those connection details it will also use certificates to read that data back and it may not be able to so you said we need to export certificates from here manually yeah so um as the host name changes for the central node we will need to issue certificates from the qmc as well as back them up manually um that is because of course the the host name itself changes if if it if the entire machine including the host name would be migrated to a to a new physical platform let's say uh but the host name is kept you only need to back up the certificates manually because the issuing authority as we will see in a moment won't change it will still reference the uh the central node the central node always generates and signs all certificates in a click Send site and the number one consideration here here is that you have to reference the fully qualified domain name of the new machine and our Target machine will be send server to so let's go ahead and do that and these these certificates of course are not uh the ones used for proxy communication so the these are not a certificates that you present to your users via the browser but rather the certificates that click Send Enterprise uses to communicate within its own Services a certificate password is recommended although not mandatory we will go ahead and leave this plan for now however pay Extreme Caution to this option during a site migration it's mandatory to include the secret key in the uh trusted route certificate otherwise the certificate chain uh will be broken and the new system will not be able to validate its own certificates properly now there's a couple format options there do we need to export them in both formats or just which one is more important only Windows format is uh is required and and in fact it's the recommended format this PM format is a Linux compatible format for example if you if you want TLS security between your central node and a postgress database like I have set up you can use those to protect that traffic as well and I see there's a path there telling us where we're going to export them to yep and this path by the way is local to the central node so uh let's assume I'm browsing to this address from a client computer of course this path will not be present on the client computer we'll have to remote in just like we are now into the server and grab it from there I've copied the uh the path I've exported our certificates let's go take a look at what was generated you can see this uh there is a new folder here matching what we've introduced in the machine name for the certificate export and it contains three certificates the server the route and the client all three must be transferred over to the new machine so uh what I typically like to do is copy them to a different location uh I have a shared folder here set up for this exercise so we'll paste them in here and leave them as that now the next step would be to uh back to bump manually and here's a way that you can run Microsoft Management console as a different user in this particular case domain back/ QV service so you can navigate to see Windows system 32 find your MMC executable shift right click to run as a different user so we'll add two different snap-ins one for the current user account and since we logged in with account that runs the Cent services that will give us access to the right certificate store the other one is the computer account and this is where your trusted root certificate will be present so let's start with the local account first and these are your machines trusted root certificates your list may be uh much longer depending on what other trusted certification authorities your company trusts but the uh certificate that is autogenerated by click sense will typically have the fully qualified domain name of the machine it was generated on and that CA format we'll go ahead and Export we absolutely must export the private key as well otherwise this certificate will not be uh useful to us on the new machine we also should export all extended properties to make sure that all current properties in the certificate are transferred over to the new machine we'll go ahead and set up password so that it lets us export is this going to export in the same format as the manual export yes uh by default when you're exporting things from window own certificate store they will always be in uh pfx format or Windows format as detailed in the qmc and Mario what's like the real purpose for exporting manually and through the MMC um the original certificates uh that we are backing up right now are essentially a worst case scenario recovery tool the new certificates that we've issued through the qmc will be used on the new server again always export the private key always export extended properties and your click uh client certificate as well the same procedure always exporting the private key always exporting extended properties and is this a password that you're kind of making up now or is it a pre-existing password it's not an existing password it's just a password that Windows uses to sort of encrypt the certificate and it is required when you repport it okay okay so now with our certificates backed up it is time to take a note of the service cluster location this one we can close already so let's go back to our qmc your service cluster is defined here so if this path changes at all before starting up the uh services on the new machine we'll have to make sure that we've uh verified these settings and have updated the paths um using a a small configuration tool that I'll I'll show you a little later but for the purposes of this migration the new machine has access to that share yes and that is of course part of your planning as well um you should always test access to all shared resources uh between these machines so we've got our service cluster we've got our certificates backed up now which Step actually makes the backup everything in the qmc here all right so for that we'll need to actually back up the uh repository database that is running in the background that is based on postgress and the details on how to do that are available on on the help site here we can see the step-by-step instructions and we start by stopping the uh click Send Services except accept the cence repository database as we will access this to uh take that backup you will need to open a command prompt with administrative privileges so copy and paste your um command line from the help site however pay close attention to the uh destination path for your backup ours will be slightly different and we can talk a little bit about what we are typing here so PG dump is a built-in command from your postgress utilities that is available uh with any installation of Click sense Enterprise we're defining the Local Host uh as the target host if you happen to be running on a dedicated postgress Host this would have to change uh your Port your uh main user that we'll be using to uh connect with this is the user that we Define the super user password for during setup um and that will be the password that is required and some options for taking the backup we're using a binary file um we are specifying that the file should be a tarball as we can uh see thear um extension and the last uh bit of information here is the database that we wish to uh back up so this is not optional um we need to specify qsr as this is how the database is defined internally and before we do all that we'll need to make sure that our services are stopped so again we'll be relying on everything but the uh repository database to be stopped a quick way to do that is to stop the repository service which will in turn stop every other service that is dependent upon it with the exception of the service dispatcher which will have to stop manually um your logging database is unaffected by this process of great so everything stopped so now we run the command and that creates the backup of our repository right that's right Troy a small tip I have for you when uh running the backup command it's always a good idea to use the dash V for verbose flag this will have a more complete output on your screen about what's going on and if something should go wrong we'll be able to catch that before we actually start the services and correct the problem that's a great tip it should be the first argument it's asking for our password this is again the super user password that you've defined during installation and that's it our backup should be ready and here we can see the backup file is present and ready to go so we have our certificates we have along with all the apps and app objects users user attributes your data connections your tasks however please keep in mind that there are special considerations for changing domains changing the host name and moving your service cluster to A New Path but basically this will put put everything back the way it was same apps same list of users same active directory same tasks it'll just rebuild it in a new environment with the same database that's correct Mario what happens if we missed one of those steps like with the certificates forgot to include the private key or something like that so let's say the private key is missing most likely the the repository service will not be able to start fully and if it is able to start it won't be able to decrypt all the data in the database I mean it's just a checkbox so it feels like a human is pretty possible there would it actually tell you that you forgot to include the private key yes there is a log file in the following location so we'll have to go to your log folder under program data we'll search for the repository trace and the security uh repository log file will contain information about your secrets and your crypto key access okay let's say your click client certificate was not exported with its private key the ultimate effect is that all your data connection passwords won't be deed rtable anymore so let's say you have 15 data connections that use passworded connection string you will have 30 error messages in this in this particular log is there anything left to back up or do we move on to the importing process of the new system there is one more item that I uh recommend backing up always and that does require stopping all services including the repository database and that is the contents of your program data folder and contains a raw copy of your database and it also includes your local logs and a couple of other items like your host config it's good to have in case something goes wrong and the original system is no longer available and must be rebuilt this is one of the items that you'll use to uh to rebuild this system and that's just a straight up copy of the entire click folder yeah one more item please consider your custom connector packages uh either from click or a third- party provider if you use uh obbc you will need to recreate these connections on the new server um preferably with the same names so that you do not have to then go in and edit any of your click scripts um or modify your existing data Connections in the new environment but aside from that we now have a full site backup uh and we're ready to bring this content over to the new machine where click sense September 2020 release is already installed we have decided to create a new cluster with its own database so there is a postgress instance on that machine as well so that we can import the data after so as you can see here we have all services installed and running under the same uh service account as uh send server one so what we're looking at right now is just out of a box install everything's default pretty much fresh and ready to go this is the first time I'm logging in it's ready to be licensed there is nothing going on here first things first we will start on the destination Machine by importing those certificates I recommend that we start this uh process by stopping all services do we leave the database running yes we'll we'll take everything down just to make sure that no ports in news nothing that can uh prevent us from from doing what we need to do here so here I'm accessing the um the shared location where we've backed up uh all of our content let's go ahead and import these certificates first find your MMC executable now of course when you set up a new environment or a new server to receive data from an old server you will have certificates we'll need to get rid of those as we are importing certificates from an existing environment and we don't want there to be any clashes inside a certificate store or any duplicate values that may confuse the services when starting up that makes sense you want one set of certificates nothing that might conflict when you're importing certificates are there any Windows security rules that might cause problems when you're importing absolutely good great point I'm glad you uh you brought that up um new security policies have kicked into place that may prevent services from starting up completely I'll I'll show exactly what I'm talking about um there is a local security policy under Security Options the policy that we need to look for is under the uh system cryptography setting um and it's uh this one system cryptography for strong key protection for user user key storage on the computer if this policy is set up to any of these two last options here on the menu you will be first prompted to assign a new password that has nothing to do with the password that you've used to encrypt the certificates when backing up windows will require you to enter every time you access it should be set to that first option or not set at all as I you're saying not set at all just to make sure that you can at least import the certificate and mind you that option has to be disabled only during the short period of time that takes you to imported certificates you can enable it on the server again after the certificates were imported uh because that policy governs only new certificates it is not retroactive that's another great tip okay back to importing so uh let's start by importing The Trusted D certificate under the local computer account uh and as I mentioned before since this setup uh was started albeit in a blank State um the certificates have been generated for the current empty install so we'll we'll need to get rid of these uh to make sure that there are no duplicates no backup necessary I'm just going to go ahead and delete these now this uh new certificate that we didn't see on the origin machine click service cluster certificate this is only useful to multicloud deployment sites where you are Distributing applications to uh click cloud services uh it is not the case here so I'm I'm ignoring that certificate as well the other certificate that we need to get rid of is uh click client that's all gone we'll go ahead and import them in the right order right click here all tasks and import the correct store will already be pre-selected and cannot be changed and we'll go ahead and navigate to our backup location as this is a complete swap of a central node to a new host name Will import the certificates that were manually exported from the original machine I just want to point out that importing from either certificate backup is correct but click recommends using the qmc export with the correctly find name and then we'll go ahead and import the server certificate into the personal store Mark the key as exportable so that uh you can then take another backup that includes all properties and since we did not Define a password during uh the export from the qmc we won't have to specify one here and the last one is going to be our click client certificate if you don't see uh the certificate that you're looking for uh just drop down and select all files and select your client seert but here is another option that users must pay Extreme Caution to this first option directly relates to that local security policy that we saw about uh strongly protecting local client keys please do not check this as this will then prompt for a password every time the certificate is being accessed test and of course we want to make sure that the private key is uh exportable and we include all the extended properties and we'll leave it in the personal store and that's it okay so all the certificates are imported what's next yep certificates are in place what's next database operations first drop the current empty database you can drop it via PG admin if you have this installed or available via a different service or you can do it by the command line the uh steps for the command line are detailed in The Help uh however the steps for PG admin are not but of course it's a it's a more visually appealing tool and easier to use so let's uh go ahead and take a look at that let's set up a new connection our current server is 1015 the standard Port where the postgress repository is running is Port 4432 the username and the password will be the one defined during setup there we go there we go as you can see you have uh several databases in this uh database engine um so this is uh an easy step as you can see we'll we'll have our schema and our tables this is where click sense stores all of your data or you can use PG admin to manage your database to take backups and restore we'll have to drop the table of course you can do that by delete drop alternatively we can use the instructions from the help site we'll go ahead and copy this content as we've copied it from the help site and drop it in here uh again we are connecting uh locally to this current database host with the port and this username and we are trying to drop the qsr database no other tasks running in parallel for example a backup task running on an automated script when you are trying to do this just rightclick and disconnect the database wow so let's go ahead and do that now we've enter the super user password and there we go so now going back to PG admin I'll just go ahead and refresh and you can see qsr is gone we'll use the catb command again connecting locally to the same port with the same username we'll use the template uh zero to give us a basic schema back end however there will be no tables recreated based on this or anything else all of that is uh taken care of by the repository service during the initial startup so it seeds the database if it's empty simple as copy and paste right click to paste here is there a way to add a uh d v to add some verboseness to that so we actually see what happens as well yes I do recommend just as we used uh- V for verbose flag during the backup we will use the same during the restore so that we get an actual feed of um all the tasks that are that are being run in the background so let's go ahead and create this database apparently I cannot type my password right now and there we go so just to illustrate we'll go ahead and refresh this again and here we have a qsr database with a public schema with no tables so how do we get this data back of course you can still use uh PG admin to restore your existing file you would use the restore function and navigate to your tarball and click restore watch for the progress we'll go ahead and use the manual method um which is defined in our online help you can see here I've added a variable path to backup file I'll copy everything up to that point and then we'll go ahead and copy the path of the tarball manually so we'll paste that in here we have our qsr backup again this is currently sitting on a share Drive uh we'll go ahead and bring this locally I've created a folder here beforeand called qsr backup and a quick way to grab in paths from from existing files is again to shift right click and there is this wonderful option here cop as path that already includes double quotes um so that if this sits in a folder structure with spaces in it it would work just by copying it and pasting it here as we can see it is recommended that you add- V for verbose output to this command as well this is well we we'll spot any any problems during the import procedure before that let's go ahead and make sure that we're disconnected from the database so that we can use it and we are ready to hit enter yeah I love that verbose tip it's so satisfying seeing streams of text scroll across command prompt indeed you you very seldom do something in the in the command prompt that doesn't offer an immediate output yeah there we go creating indexes constraints uh adding foreign key constraints adding the data um if there were any problems here at the very end of this uh message we would see that the pro process something like process finished successfully with so and so many uh warnings we've restored so now we can verify that the uh that the data is actually there let's refresh go into our uh qsr database and look at the tables whereas previously uh just after recreating the database from template there were no tables now we have all all the content so now we have our certificates imported we have all the repository metadata U present on the new system and our Rim node is still active in talking so we'll try to stand up uh the services on the new Central node we'll make sure first that the old Central node is it fully stopped that's back on sense server one yep that back on sense server one will'll make sure that everything is stopped and it is repository database first service dispatcher second uh and the repository service a close third and we'll be monitoring the activity on This Server uh on on that server specifically just to make sure that everything is uh coming up as expected so we'll navigate to our log folder into repository and Trace and you can see here there's a bunch of new log files these are automatically rolled over upon service restart and and after reaching a certain point during the startup phase they will be automatically moved over to your archived logs folder location if everything goes fine you will eventually see these files magically disappear I am comfortable to start bringing up other services such as the engine proxy your printing your scheduler if you see this in your logs there is typically one root cause the repository service at this point since we have aha on purpose missed the step won't get past this point the internal host name for the central node inside the click sense repository does not match your machine name nor the machine name for which the certificates were reissued soara I'm a little confused which Step did we miss uh we missed the old important bootstrap step this is a command that you need to run on the new Central node that will automatically update the host name values for the new Central node in inside the database and also recreate any certificates that are that are necessary based on the trusted root certificate available on the machine so let's go ahead and run through this process now we'll have to put it in bootstrap mode run it in Standalone so that it runs through this special mode and then restore host name is the special command that we need to use to make sure that the internal values are updated of course need to make sure that the repository service is stopped and every other service is stopped on the machine other than the repository database the service dispatcher does not crossing my fingers so this will uh recreate any certificates that are missing and also bind them to the appropriate Services um and once it enters main startup phase uh it will exit out again and at that point uh we should be comfortable that uh the database has been one fully restored and two properly configured to operate on this new machine so once again in this time for good let's uh start things up and we'll once more monitor the uh startup procedure to make sure that everything is running smoothly and real quick again what is the order for starting at the services service dispatcher first repository database um repository service let that run for a few seconds uh make sure that things are starting up correctly by looking at the file and then um everything else of course the logging uh the login uh service database I have to start that before anything else as well um and as this one is independent from all other services um but the my preferred order afterwards after repository service is fully running is to go engine proxy schedular printing here we go uh nice and clean folder here means all of our previously generated log files were successfully archived that means our service cluster is fully accessible by the new system we have the right permission set up for the uh service user so are we at that glob ious points where we get to check the Hub yes sir yes we are Let's uh log into the qmc first and we have a platform let's check the note status again these are still talking to each other still 5x5 on the rim node it looks just like it did so the final test dealing with existing data connections that have credentials associated with them one quick test uh is to just run one of those uh reload tasks and see if we can get the app again uh so for that let's just go to the hub we're authenticated we can see our applications let's go look at our postgress SQL data test now this is an app just to sort of test that data connection you've set up that's correct that's correct it's currently set up to reload once a day uh but we can trigger this manually real quick just to uh make sure that the uh the time stamp is updated let's go ahead and reload that task and make sure it's still running it already finished latest data reload 12:35 so everything seems to be working so far there are a number of other considerations that we didn't touch on today the bulk of the time taken here is prep work and then actually uh backing up your content we recommend a maintenance window for any such procedure are there any resources available to make this process easier here's some uh documentation on how to uh set up click sense from scratch using silent install method that's for setting up in terms of automating the backing up and restoring of the environment the simple Powershell script that triggers the uh PG backup and the p your restore command on the target server would be enough here's a great place to start for what happens next this will set you up with great activities to run on a daily weekly monthly quarterly and yearly basis of but I highly encourage you all sense admins to start here and familiarize yourself with this content and with that that was it for me thank you all very much for joining us today Troy now it's time for Q&A go ahead and place your questions there anything questions you might have in the Q&A panel on the the left side of your on 24 console Mario which question would you like to address first well Troy there's a a very interesting question here about update sequence um somebody's asking when upgrading a multinode click sense environment can it go straight to the latest update or is it best practice to install update sequentially um I'm not sure if everyone is aware uh about this but all of our updates are cumulative um so if you jump between two major feature releases uh let's say between April and September you will also get all the content that was delivered in the June release uh which was skipped so there is no need to do sequential updates um however be careful uh with how big a jump you're making between between versions um for the very very old versions there are some special considerations however when moving between modern versions post um feature releases in in 2018 you can just jump over to the uh to the latest um I would suggest that the bigger jump you make uh the more testing and validation you should do to make sure that everything in the system works fine including applications and their objects as we um we have made quite a few changes in recent releases and added new content uh but other than that yep you can uh you can jump straight to the uh to the latest the same applies for Patches by the way so um you can go from patch one to patch five knowing that everything on two three and four is also included okay next one all right uh let's see what uh special considerations are there if we're migrating from a physical server to an AWS environment um well pretty much all the networking considerations they have in a Windows environment uh physical on Prem except you will have to deal with aws's networking Services Network load balancer Etc and make sure that the um access route that users are going to take to reach your system is uh fully open and configured to accept uh connections for click sense um if uh if there are any any doubts about how to how to achieve this or something isn't working as expected feel free to contact click support would love to take a look and help you out all right uh let's see a bunch of really interesting questions here um yeah someone uh is mentioning that there's a new certificate that showed up during the uh presentation um that we haven't seen before this is called click sense uh uh click service cluster sorry and um it's a brand new certificate introduced in the September 2020 release and this is used for um multicloud environments that have a Windows on Prem installation and want to distribute apps to uh a kubernetes uh install or to our own click cloud services um and just as uh the rest of the certificates this one also has to be backed up during a migration process or during your normal backup procedures um and will have to be restored as well in the case of a migration um the uh help uh click help site has been updated to reflect this step as well so if you follow those procedures you'll be fully covered on certificate backup and restore and uh another short question um what about using block storage um if you're referring to the service cluster um using block storage there is not advised as will will need um NTFS file policies to be uh to be enforced which means the file share uh hosting your service cluster should be hosted by a Windows box um just acting as a file server so that you can always uh keep permissions under control and it's easier to troubleshoot we also support um certain type of Nas configurations provided that the uh permission layer of NTFS or SMB protocol is um the very latest one used and there are no restrictions to file access um the reason why this is a new change is that in the old days uh smaller capacity Nas boxes for example would have restrictions such as file listing um so this is something to keep in mind but for blog storage specifically um we don't currently support this as far as I'm aware um because it does not support the NTFS permission layer and our product relies on it to um to control access to files however um I do recommend that in um Cloud providers in your own virtual cloud provider you you try to deploy um instances with flash storage wherever um file hosting is uh is going to be placed and and that is especially important for the server that hosts your service cluster um flashbay storage is going to be faster than uh then hard drives and it will help performance as well as uh access times to uh to smaller apps um and someone is asking if uh the process that we've walked through today is the same as for migrating click viiew server um and the short answer is no um these two procedur are of course very very different uh because the products are very very different um if you want to know more about how to migrate a click viw server correctly uh from machine to machine please refer to our online help and knowledge based articles that are present on click community and again for any issues uh that may occur during that process contact click support uh yeah there's there's a couple here uh that have to do with uh changes to apps after such migration U there's one asking about apps needing to be imported individually um of course not um in the presentation we saw a migration process that did not involve moving the service cluster over to a new location um however the service cluster hosts all of your apps and and other app metadata uh that is not present in in the database so by directing your new environment to an existing file share that hosts the service cluster or by moving those files across you are essentially moving all of your apps over so you won't have to import them into the qmc one by one uh and the other question question is um for the operations Monitor and other monitoring apps available in in the system can they be configured to read from the previous environment logs as well as the new ones and the answer is yes of course um all you have to do is go into your um into your system enable a security rule so that one of your rout admins or one of your administrators can see all of the um backend rest connections um and the uh the folder connections and just create one for the uh for the new location how however for the old logs um the old logs will also be present in your qvds as they get reloaded on the old system qvds would have been generated so if you move those across to a location where the uh current monitoring gaps can can access them that data will still be present so you're not going to have a clean split between um uh between data in these monitoring apps unless you get rid of the old qvds or never import them let's say uh and just reload on the new environment from scratch that would then only show you the the new environment data okay Mario we have time for one last question all right so um somebody is asking here about upgrading from a old version of clicks and say 3.2 to the latest which would be September 2020 um as well as upgrading their Hardware at the same time this um this type of migration uh I would raise quite a lot of caution here this is a huge jump and many backend modifications have happened in the product since uh click sense 3.2 or the numbered versions um there are uh very detailed articles on on click support and and click community on how to do this correctly it involves jumping to a intermediary version um somewhere slightly more modern than click sense 3.2 but not quite as recent as one of the 20120 releases um and then from that point onward you can jump over to any of the uh the latest releases uh however as advis during the presentation I would first tackle the product upgrade validate that everything is still working working as expected correct any issues that you see there and then once happy with the performance of uh of the existing system at least on the new version considering any hardware limitations um I I would then um start planning the migration to to new hardware all right so that was uh that was it for today thank you all very much for joining us and submitting your questions uh this has been a blast and looking forward to the next one okay great thank you everyone we hope you enjoyed this session and and thank you to Mario for presenting we appreciate getting experts like Mario to share with us here's our legal disclaimer and thank you once again have a great rest of your day [Music]

Channel: Qlik Support

Views: 1,256

Keywords: Support Techspert Thursdays, yt:cc=on, STT, Video, Support, Qlik Support, Troubleshooting, Best Practices, Qlik, Support Expert, Thursday, Webinar, Qlik Webinar, error, fix, recommendation Qlik Sense, Enterprise, Upgrade, Migrate, environment, PGadmin, backup, Back up, back-up, recovery, restore, certificate, connection, QVD, users, QMC, tasks, app, QVF, hub

Id: -ZZ6nNILPJo

Length: 46min 3sec (2763 seconds)

Published: Fri Oct 13 2023