Jump to content


  • Content count

  • Joined

  • Last visited

  • Days Won


OlgerDiekstra last won the day on March 19

OlgerDiekstra had the most liked content!

Community Reputation

12 Good

About OlgerDiekstra

  • Rank

Profile Information

  • Title
    FileMaker Developer
  • Gender
  • Location
    Gold Coast, Australia

FileMaker Experience

  • Skill Level
  • FM Application
    16 Advanced

Platform Environment

  • OS Platform
  • OS Version

FileMaker Partner

  • Certification

Recent Profile Visitors

7,246 profile views
  1. OlgerDiekstra

    Connection problem with FMS 16

    Your local and external ports are hardly ever the same. The local ports are managed by your server. The external ports are managed by your internet router. On your router you can redirect ports to different internal servers. Ie, you can have a different server act as your webserver, while filemaker resides on a completely different machine again. From external you access these machines via the same ip address but on a different port. On the router you specify which port is redirected (forwarded) to which internal machine. Without knowing too much about the network, I would say you are trying to access your filemaker server with an external ip address when using dns. Try an nslookup and see what ipaddress is returned. Or maybe your external dns has the old ip address still listed and that needs changing.
  2. OlgerDiekstra

    FM server specification literature

    I doubt you'll find a paper that can discuss this in meaningful detail. It all depends. Number of clients is just a small factor. Scheduled scripts can impact performance, schema design of the database(s), Webdirect, PSOS scripts, nowadays PDF creation on the server (as in how often these are triggered), and other services running on the server can all have an impact. It really depends very much on the DB. Future developments can also have a detrimental impact on performance. Which is why I suggested baselining your current environment, and decide based on that. I've got a (5yo) Win2k8 server with 16GB RAM, 8cores serving 12-15 office LAN users, which works well in terms of performance. We're now also starting to deploy iPads, which will eventually end up around 40 or so. Be very interesting to see how that goes. But as these iPads aren't deployed with a big bang, I've told the business we either see how things go, or we upgrade now just to be sure. But that's a big expense they're not keen on. So we'll just tread carefully.
  3. OlgerDiekstra

    FM server specification literature

    Review current performance history in terms of CPU usage, memory usage and network usage. If you haven't got historical performance info, start collecting now and do this for an acceptable time (ie a month if you're keen to move soon). That will give you insight in how many spikes there are in terms of CPU and network usage in your current setup. Then liaise with the business what you want to do and why, present the data as evidence that it's a sensible move. Get the business to agree and accept that there's a risk, in which case if it all goes sour you've covered yourself. If your current m4.xlarge is tapdancing along, I see no reason why a t2.xlarge wouldn't suit as well considering the specs, but I haven't used AWS server instances.
  4. OlgerDiekstra

    Realtime Developer Intelligence

    That would make most tools 'realtime'. They are all opening/importing the DDR xml files. They all allow you to click on something or search for something so they are analysing/processing. FMPerception opens and processes the DDR xml no different from the other tools. It's really just a pretty interface to the DDR xml that lists the content smartly. When discussing performing an analysis realtime, you have to consider the source that provides the data to analyse. FMPerception can't do anything without the data. Performing realtime analysis means updating the data in (near) realtime. Which currently can't be done. So concluding that FMPerception isn't a realtime tool is accurate imho.
  5. OlgerDiekstra


    I have to disagree here. Upload AND download speed play a role, but you can have a synchronous 1GB link to your ISP and still not get more 1Mbps throughput. Interconnects, distance, all play a role. Ping response times can be deceiving, indicating a good link when using small packets, but when connecting with anything else it may fail. In fact, pings will mask routing errors, giving the appearance that there's a functional path whilst in reality there's not. Router hops is mostly insignificant. You can have a link with 5 hops and high latency, or a link with 30 hops and low latency. The latter vastly outperforming the former. Satelite links are a good example, often showing 800ms+ latency with just a few hops. We've used a 10M/10M link for years at the place where I work, only recently (couple months ago) upgrading to 30/30. My home connection is ADSL2 with 7M/384k. Works fine when using an iPad FileMaker app connected via a VPN back to the office. Of course, it's easy to saturate the uplink. And Skype can be troublesome, but it still works.
  6. OlgerDiekstra


    Exposing a v13 FM server to the internet may not be smartest approach. There are other, more secure ways to provide access. Teamviewer can be used to remote in to a dedicated workstation and gain access to the FM server. Benefit is that a dynamic IP address is a non-issue. There are others such as Logmein that provide the same sort of functionality. Another benefit of this approach is that the FileMaker data never leaves the premises. Alternatively, a VPN can be used which can either terminate on the clients internet gateway/router or on the FM server. You would need either a static IP address or use dynamic DNS. DynDns.org is, as Ocean West indicated, a well known provider, but there are heaps more.
  7. OlgerDiekstra

    Realtime Developer Intelligence

    Whilst FMPerception is a great tool (I use it myself) I agree that it's not a realtime thing. Geist's blog states that too: https://docs.geistinteractive.com/article/62-quickly-running-ddrs in the opening sentence. You may be able to automate it, but that doesn't make it realtime. If it were realtime, I'd be able to make a change in a DB and see the results in FMPerception instantly. That doesn't happen unless I trigger a new DDR export. I wouldn't want to have a new DDR report generated every time I made a change though. That'd be a real resource hog.
  8. OlgerDiekstra

    One Master Product Catalogue - Many POS Different Locations

    Sure, FileMaker can drive something like this. You could use its new API to pull and push requests (using JSON for instance). Create vendor logins and track for each products which vendor is the owner. A cloud based service might be the way to go, but in the end the cloud server is just a fm server someplace. The thing to consider when self hosting is can you (do you want to) ensure it's uptime? It may not be an issue, if all vendors access the system only during certain hours within the same timezone, you only need to worry about 9 to 5 or something. But if they are across many timezones, things get more complicated. 10,000 items is nothing. Things to consider though is performance across WAN/Internet. Depending on how well the DB is designed, pulling a list from the central DB could take a while. But you could create an automated after hrs sync and only sync changed items. Realisticly, how often do products change? Probably not daily, certainly not all products.
  9. OlgerDiekstra

    Lookup import advice

    I have various tables with a million records (maybe more, haven't looked for a while). Lots of records doesn't imply things get slower. How the schema is organized/structured does.
  10. OlgerDiekstra

    FMS Backup Strategy Best Practices

    I manage a Windows 2K8 R2 FM16 server. Full backups are done nightly (daily: Mo/Tu/We/Th, weekly on Fr, none on Sa/Su as no data gets changed then), with progressive enabled. C: drive is the OS drive and holds the Applications. D: drive is where the databases live. E: drive is where the backups are stored. C:, D: and E: are partitions on a RAID drive. A batch script copies backups nightly to a NAS. The NAS copies FM backups (along with other backups) to another NAS. The other NAS backs up to AWS. Everything is fully automated and requires no human intervention, aside from the checks. If backups fail, I get alerted. Generally, no news is good news, but the system is still regularly checked.
  11. Show/hide menu bar.
  12. OlgerDiekstra

    Custom Function for an article A or An

    You can make the If( left( $term; 1 ) = "e"; "an "; "a " ) into a custom function: function repl_term( _term ) Case( Left( _term; 1 ) = "e"; "an "; Left( _term; 1 ) = "a"; "an "; Left( _term; 1 ) = "i"; "an "; Left( _term; 1 ) = "o"; "an "; Left( _term; 1 ) = "u"; "an "; "a " ) & _term And then do Substitute( $text; "<<subst1>>"; repl_term( $term ) )
  13. OlgerDiekstra

    System Script user account

    Have you tried "putting" the test.txt file somewhere outside the program files folder? Ie, in the user folder of Ben? Windows (as of v8 I think) gets pretty cranky when you tried to write to the program files folder. I would say it's a permissions problem.
  14. OlgerDiekstra

    Custom Function for an article A or An

    You're making it more difficult than it needs to be. When you substitute <<subst1>> or <<subst2>> in your sentence: "<<First Name>> <<Last Name>>'s reading performance was found to be at <<a or an>> <<subst1>>, while <<his or her>> math calculation skills were found to be at <<subst2>> " you can determine <<a or an>> when you substitute. In which case you don't have <<a or an>> in your text, just <<subst1>>. So your user selects "eight" from a dropdown list, and right before you substitute in your CF you determine whether eight requires "a" or "an". The text then becomes: "<<First Name>> <<Last Name>>'s reading performance was found to be at <<subst1>>, while <<his or her>> math calculation skills were found to be at <<subst2>> " Something like: Substitute( $text; "<<subst1>>"; If( left( $term; 1 ) = "e"; "an "; "a " ) & $term )
  15. OlgerDiekstra

    Custom Function for an article A or An

    I think you are looking at this the wrong way. In your example it makes no sense to substitute only <<a or an>> as they would never change based on the text. The substitution would always be the same. So, for your example to make sense, the words following <<a or an>> (ie, eight and fifth) would also have to be substituted. In which case it makes much more sense to do something like this: "The client's reading performance was found to be at <<subst1>> grade level, while his math calculation skills were found to be at <<subst2>> first grade level. " And substitute <<subst1>> with "an eight" and <<subst2>> with "a fifth". Whilst you are substituting you could check whether the word starts with a vowel or not and include "a " or "an " depending on whatever rule you define.

Important Information

By using this site, you agree to our Terms of Use.