Our only Linux server is a headless machine in a remote data center.
I’m taken aback to discover that there seems to be no way to install Servoy server via the shell.
The docs suggest that I install on a server that does have X windows and then copy the install across.
The only way I can think of is to mount a version of Linux with Active Directory such as Knoppix on my XP workstation. The problem is, the distro sites aren’t very forthcoming about what they are loaded with, and I don’t have time for trial and error.
Can someone please recommend a distro that has suitable versions of Java etc so the install will run? Or if I’m on the wrong track, any alternative suggestion would be most welcome.
If you have a Mac sitting nearby you can install on that and just move the folder to the linux server. Another easy and quick way is to use VMWare, download a preconfigiured VM Image with an easy linux like Ubuntu, install and move the folder to the linux server. If you don’t have VMWare installed you can also download a “Live CD” from Ubuntu website and start your PC with that, it doesn’t require any installation and you can ftp the folder to the server and reboot in Windows in minutes.
Keep in mind that you will need to install the official Java from Sun to run servoy but I think that the default third party java distributed with ubuntu should be enough to run the installer.
I’ll try Ubuntu - haven’t had much luck with Knoppix…
It really shouldn’t be this difficult - I would have expected a terminal based installation option - not the most user-friendly introduction to the product
Well, what you are doing is not exactly an introduction step… you are setting up a production/test server. If you need to setup a development installation you also need to run Servoy Developer and then X-Windows is needed anyway but I agree that a shell script to install using the CLI on unix systems would be a welcome addition.
ran the Servoy jar (you have to right-click on it in the file-manager and select the runtime - double-clicking the jar simply drills down into the jar file)
copied the Servoy directory from the Linux /tmp file to the Windows hard-disk
I’m copying this across to our headless server. Hope this’ll work - it’s certainly been more time consuming than I’d like.
Our scenario (develop on Windows, with the data on a remote Linux server) must be a common one - surely there should be a shell based install option for the server component?
In the meantime, why couldn’t Servoy make the server install directory available as a download? It’s not as though there are very many options with the install wizard?
In order to install on headless you have two options:
On finish panel of install you can save an xml for automated install; then use it as: java -jar servoy_installer.jar myxml.xml
There is a new feature in izpack, console mode. You can start it by: “java -jar servoy_installer.jar -console” and install in console mode. Unfortunately this doesn’t have full support in izpack, for example will install only the default packages. Also this feature is included in latest release of Servoy only (4.1.x), when we updated to latest izpack. (so option1 is probably the way to go)
Do note Servoy server does not start any database!
Struggling to figure out how to start the database. The docs suggest running developer using a shell script that doesn’t exist in this version, and the install doesn’t seem to have created a service for iAnywhere.
I’m struggling to figure out the startup options required to make the db work with Servoy. Can some kind soul please supply the shell command?
I’m sure the would be useful to others as well as myself, as it seems to be undocumented?
Do note Servoy server does not start any database!
Struggling to figure out how to start the database. The docs suggest running developer using a shell script that doesn’t exist in this version, and the install doesn’t seem to have created a service for iAnywhere.
I’m struggling to figure out the startup options required to make the db work with Servoy. Can some kind soul please supply the shell command?
I’m sure the would be useful to others as well as myself, as it seems to be undocumented?
Geoff
Many times is useful to search the forum for common issues (like this one) as they were probably already answered somewhere. See , for example this post: viewtopic.php?f=5&t=12447#p63043 .
It is surely unfair to criticise me for not spotting a post relating to Windows bat files when I’m trying to solve a Linux problem (see title of thread). And in any case the post doesn’t seem to be relevant, as the sybase.config file it refers to doesn’t exist on this Linux install.
I have searched the forum quite carefully, and if the answer is here, it’s hard to find. And as I say, this scenario seems to have been omitted from the docs. I don’t know Servoy and I don’t know iAnywhere and I’m looking at around 2000 pages of docs. So to save me an hour or two of frustration, it would be much appreciated if you could give me some more practical help with this.
It is surely unfair to criticise me for not spotting a post relating to Windows bat files when I’m trying to solve a Linux problem (see title of thread). And in any case the post doesn’t seem to be relevant, as the sybase.config file it refers to doesn’t exist on this Linux install.
I have searched the forum quite carefully, and if the answer is here, it’s hard to find. And as I say, this scenario seems to have been omitted from the docs. I don’t know Servoy and I don’t know iAnywhere and I’m looking at around 2000 pages of docs. So to save me an hour or two of frustration, it would be much appreciated if you could give me some more practical help with this.
Thanks
Geoff
Hi Geoff,
That was merely a suggestion, didn’t want to criticise you. The sybase.config file should be located in application_server/sybase_db. So, the shell script should be:
(Not sure how I missed the servoy.config file - must have had a brain-fart I did try adapting the bat file, but it was beyond my limited sysadmin skills…)
Seem to be functional at last: server has started OK and developer is talking to my remote sybase via SSH tunnelling.
But I’d appreciate some advice on setup. With developer on my local windows machine and server on the remote Linux box I seem to have 3 options:
Devolpment Metadata in the default db on the Windows box. But then I have to transfer solutions to the linux box to deploy, and access to developer over our slow VPN when away from the office would be painful. So unless development with remote metadata will be very slow over ADSL, I prefer to avoid this one.
All metadata on the remote box, in Sybase. This might slow down development a little on the Windows box, but enables me to work from my laptop. I have to worry about backing up Sybase as well as Postgres
Move all metadata to Postgres on the remote box. I have a tested Postgres backup system, so this simplifies life and saves some memory.
If you can point me at documentation on how to migrate the metadata, option 3) would be my preferred solution, unless there’s a problem I’m missing. What do you think?
Not sure I fully understand this but here are some advices (hope they help). First, if there is only one developer(one person) you should develop the solution locally, only using localhost. Then you can export from developer and import the solution on your remote server (test or production server ?). You can also checkout from remote server, and use team provider for update/commit but import is better and clearer option to deploy. If you want to switch to postgresql: go to servoy-admin page of your headless server (http://server:8080/servoy-admin), there is a servers page where you can add/remove/edit servers. A database of a server must be created outside servoy, upfront. The metadata is located in server repository_server. Modify the connection so that it points to an empty postgresql database (maybe add another test server first to get the connection data right). After that you must restart application server with parameter: -upgradeRepository which will create all our internal databases in postgresql database. Restart again (without -upgradeRepository this time), and you will have a running application server on postgresql repository. You can also add/edit other postgresql servers as is not advised to use the repository_server for solution data. Hope this helps.
My only problem with the solution you recommend (developing on localhost) is that it would be hard for me to work with Developer if I was away from the office - for security reasons we don’t open any ports and our remote access is a screen-scraper and not very viable for serious work.
Is it really such a bad idea to have the development metadata on a remote server? I’m trying it now on the sample crm solution and it seems to be working fine.
Don’t think I understand the set up/configuration. You have an office pc and a headless server, or is there something more ? If you work with developer on localhost what ports/remote access you need ?