tode / pharo errors
Locked
 
|
As per Dale's advice im starting a new thread on this one. I run "installServer seaside 3.2.8GsDevKit script: create_gs_server_image -d /home/gemstone/GsDevKit_home/shared/pharo gsDevKitCommandLine path: /home/gemstone/GsDevKit_home/shared/bin/create_gs_server_image ================= create gsDevKitCommandLine.image image using /home/gemstone/GsDevKit_home/sys/default/pharo/gsDevKitCommandLineLoad.st ++ readlink /home/gemstone/GsDevKit_home/shared/pharo/pharo + DIR= + DIR=/home/gemstone/GsDevKit_home/shared/pharo/pharo ++ dirname /home/gemstone/GsDevKit_home/shared/pharo/pharo + DIR=/home/gemstone/GsDevKit_home/shared/pharo + cd /home/gemstone/GsDevKit_home/shared/pharo ++ pwd + DIR=/home/gemstone/GsDevKit_home/shared/pharo + cd - + set -f + /home/gemstone/GsDevKit_home/shared/pharo/pharo-vm/pharo --nodisplay /home/gemstone/GsDevKit_home/shared/pharo/Pharo.image save gsDevKitTmp /home/gemstone/GsDevKit_home/shared/pharo/pharo: line 12: /home/gemstone/GsDevKit_home/shared/pharo/pharo-vm/pharo: No such file or directory gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ls -l total 36176 ....... -rwxrwxr-x 1 gemstone gemstone 1121206 May 15 16:29 pharo ...... gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ./pharo -bash: ./pharo: No such file or directory gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ chmod +x pharo gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ./pharo -bash: ./pharo: No such file or directory gemstone@ubuntu:~$ sudo add-apt-repository ppa:pharo/stable [sudo] password for gemstone: sudo: add-apt-repository: command not found I previously installed the 32bit libraries, so everything should be present. So im confused as to what is wrong. _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
So after re-running the 32bit lib install a 2nd time, I was able to get the pharo part of the installServer to complete, but now I get a new error: Also, on the chance that something got corrupted during my attempts to fix things, I deleted /home/gemstone/GsDevKit_home and started with a fresh clone. I still get the same error.GsDevKit GemStone script: stopGemstone ================= cmdLine= stopstone -i seaside DataCurator swordfish --- 10/16/2014 21:03:56.480 UTC --- stopstone[Info]: GemStone version '3.2.8' stopstone[Info]: Server 'seaside' is not running. ...finished stopStone Copy snapshot and remove tranlogs New extent for seaside Copying extent file: '/home/gemstone/GsDevKit_home/shared/downloads/products/GemStone64Bit3.2.8-x86_64.Linux/bin/extent0.seaside.dbf' Error running shell command: '/home/gemstone/GsDevKit_home/server/stones/seaside/product/bin/copydbf' with args: /home/gemstone/GsDevKit_home/shared/downloads/products/GemStone64Bit3.2.8-x86_64.Linux/bin/extent0.seaside.dbf /home/gemstone/GsDevKit_home/server/stones/seaside/extents/extent0.dbf STDOUT: ' Source file: /home/gemstone/GsDevKit_home/shared/downloads/products/GemStone64Bit3.2.8-x86_64.Linux/bin/extent0.seaside.dbf File type: extent fileId: 0 ByteOrder: Intel (LSB first) compatibilityLevel: 844 Last checkpoint written at: 08/21/2015 19:58:34 UTC. Destination file: /home/gemstone/GsDevKit_home/server/stones/seaside/extents/extent0.dbf ByteOrder: Intel (LSB first) When writing rootPage with pageId = 17 , Found current operating system clock = 1413493436 , more than 10 seconds PRIOR to stone''s startup time = 1440187046. GemStone cannot continue running in the presence of a corrupted operating system clock. write error on page 17 DBF error is not available. ' STDERR: Error: Shell command: '/home/gemstone/GsDevKit_home/server/stones/seaside/product/bin/copydbf' failed. GsDevKitNewExtentCommandLineHandler class(Object)>>error: GsDevKitNewExtentCommandLineHandler class(GsDevKitAbstractCommandLineHandler class)>>runShellCommand:args:noError: GsDevKitNewExtentCommandLineHandler(GsDevKitAbstractCommandLineHandler)>>runShellCommand:args: GsDevKitNewExtentCommandLineHandler(GsDevKitAbstractStoneCommandLineHandler)>>copySnapshotExtent: GsDevKitNewExtentCommandLineHandler>>newExtent GsDevKitNewExtentCommandLineHandler>>activate GsDevKitNewExtentCommandLineHandler class(CommandLineHandler class)>>activateWith: PharoCommandLineHandler(BasicCommandLineHandler)>>activateSubCommand: in Block: [ aCommandLinehandler activateWith: commandLine ] BlockClosure>>on:do: PharoCommandLineHandler(BasicCommandLineHandler)>>activateSubCommand: PharoCommandLineHandler(BasicCommandLineHandler)>>handleSubcommand PharoCommandLineHandler(BasicCommandLineHandler)>>handleArgument: PharoCommandLineHandler(BasicCommandLineHandler)>>activate in Block: [ self handleArgument: (self arguments ifEmpty: [ ...etc... BlockClosure>>on:do: PharoCommandLineHandler(BasicCommandLineHandler)>>activate PharoCommandLineHandler>>activate PharoCommandLineHandler class(CommandLineHandler class)>>activateWith: PharoCommandLineHandler class>>activateWith: in Block: [ super activateWith: aCommandLine ] WorldState>>runStepMethodsIn: WorldMorph>>runStepMethods WorldState>>doOneCycleNowFor: WorldState>>doOneCycleFor: WorldMorph>>doOneCycle MorphicUIManager>>spawnNewProcess in Block: [ ... BlockClosure>>newProcess in Block: [ ... On Tue, Oct 20, 2015 at 3:05 PM, Jon Paynter <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
In reply to this post by GLASS mailing list
Did you try with ldd to see if a lib is missing? See thread with subject "[Glass] GemTools on remote server" for some inspiration. On Tue, Oct 20, 2015 at 7:05 PM, Jon Paynter via Glass <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
In reply to this post by GLASS mailing list
On 10/20/2015 03:05 PM, Jon Paynter via Glass wrote: > As per Dale's advice im starting a new thread on this one. > > I run "installServer seaside 3.2.8 > > and eventually receive this error: > GsDevKit script: create_gs_server_image -d > /home/gemstone/GsDevKit_home/shared/pharo gsDevKitCommandLine > path: > /home/gemstone/GsDevKit_home/shared/bin/create_gs_server_image > ================= > create gsDevKitCommandLine.image image using > /home/gemstone/GsDevKit_home/sys/default/pharo/gsDevKitCommandLineLoad.st > ++ readlink /home/gemstone/GsDevKit_home/shared/pharo/pharo > + DIR= > + DIR=/home/gemstone/GsDevKit_home/shared/pharo/pharo > ++ dirname /home/gemstone/GsDevKit_home/shared/pharo/pharo > + DIR=/home/gemstone/GsDevKit_home/shared/pharo > + cd /home/gemstone/GsDevKit_home/shared/pharo > ++ pwd > + DIR=/home/gemstone/GsDevKit_home/shared/pharo > + cd - > + set -f > + /home/gemstone/GsDevKit_home/shared/pharo/pharo-vm/pharo --nodisplay > /home/gemstone/GsDevKit_home/shared/pharo/Pharo.image save gsDevKitTmp > /home/gemstone/GsDevKit_home/shared/pharo/pharo: line 12: > /home/gemstone/GsDevKit_home/shared/pharo/pharo-vm/pharo: No such file > or directory > > The file actually exists, but I cannot run it. So either it is > corrupt, or some dependent library is missing. > gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ls -l > total 36176 > ....... > -rwxrwxr-x 1 gemstone gemstone 1121206 May 15 16:29 pharo > ...... > gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ./pharo > -bash: ./pharo: No such file or directory > gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ chmod +x pharo > gemstone@ubuntu:~/GsDevKit_home/shared/pharo/pharo-vm$ ./pharo > -bash: ./pharo: No such file or directory /home/gemstone/GsDevKit_home/shared/pharo/pharo-vm/pharo is missing but that the image file /home/gemstone/GsDevKit_home/shared/pharo/Pharo.image is missing and in your ls above it is indeed missing ... so we are back to the issue of why is the download of the Pharo.image download failing ... The warning may be a red herring but that depends upon what you mean by "I cannot run it" ... are you refering to the fact that the you are getting the "No such file or directory" or do you mean that you are getting a different error running it differently? Since I assume that the log output for the failed downloads is gone by now ... I suggest you start again and try to include as much of the log output as possible so I can identify the erorrs ... for some reason you are having worse luck downloading the pharo image files than I have ahd ... Oh and are you using the most recent GsDevKit_home? What is the output of `git log -1`? Dale _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
In reply to this post by GLASS mailing list
YOu are getting the following error during the coydbf operation:
When writing rootPage with pageId = 17 , Found current operating system clock = 1413493436 , more than 10 seconds PRIOR to stone''s startup time = 1440187046. GemStone cannot continue running in the presence of a corrupted operating system clock. I will see if I can find out what's up with this error ... On 10/20/2015 03:54 PM, Jon Paynter via
Glass wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
The local unix guy here at the office helped me with that one. The clock on my VM was 6months behind the times, once we fixed that, the copydbf ran without errors, but I cant get the installer to pickup where it left off. So I fall back to removing $GS_HOME , re-cloning and re-running the installer. And after waiting long enough that worked.On Tue, Oct 20, 2015 at 3:59 PM, Dale Henrichs via Glass <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
So after the server install finally finished, I followed instructions to get tode client installed on my local PC here at the office. but when I tried to log in, I got a gci version mismatch error. It seems my attempt to install gemstone 3.2.8 only went partially through, but everything used the old 3.1.x version that was already installed. So I re-ran the installGemstone.sh script to install 3.2.8. but when I tried to start the stone, it gave an error about a wrong keyfile because /opt/gemstone/product still pointed to 3.1.x. but when I changed the symlink to point to the 3.2.8 install, and then i could not find seaside/system.conf. So I copied the seaside folder to the new install dir, but then gemstone crashed on startup.So I re-ran the installGemstone.sh script again. That created the folders under /opt/gemstone/product, and the usual stop/start/status commands work. And now it shows glass/seaside installed. but why does it not find tode on the server?On Tue, Oct 20, 2015 at 4:35 PM, Jon Paynter <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
On 10/21/2015 12:52 PM, Jon Paynter wrote: > So after the server install finally finished, I followed instructions > to get tode client installed on my local PC here at the office. but > when I tried to log in, I got a gci version mismatch error. It seems > my attempt to install gemstone 3.2.8 only went partially through, but > everything used the old 3.1.x version that was already installed. Well this supposition does not match with how the tode client works ... when you select a stone to connect to, the session description file for that stone is loaded and the gci file that is explicitly defined in the session description is used for the login. If no gci file is explicitly defined, then the gci file that includes the gemstone version in it's name is used ... if there is no matching gci file then you get a "external module not found" Soooo, if you are getting a version mismatch error then the most likely explanation is that the session description file the stone on your client has a different version than the session description file on the remote server ... I say "likely" because there are a number of possible alternate explanations --- without seeing the exact error message and seeing the session description files themselves (from client and server) I can only guess ... > So I re-ran the installGemstone.sh script to install 3.2.8. but when > I tried to start the stone, it gave an error about a wrong keyfile > because /opt/gemstone/product still pointed to 3.1.x. but when I > changed the symlink to point to the 3.2.8 install, and then i could > not find seaside/system.conf. So I copied the seaside folder to the > new install dir, but then gemstone crashed on startup. The fact that /opt/gemstone/product is involved makes me wonder just what system you are using ... GsDevKit_home does not put anything into /opt/gemstone/product so the fact that /opt/gemstone/product is implied implies that you are not actually using GsDevKit_home or ???? If you are using the installGemStone.sh script, you are not using GsDevKit_home ... I'm not capable of helping if you are going to mix systems like this ... or you need to be very clear about what you are trying to do by running scripts that aren't part of GsDevKit_home .... Dale _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
Ok. so I went down the wrong road here... The stone seems to be running according to the 'stones' command. next I need to start netldi with the correct port, so I stopped the stone, but I cannot get the stopNetldi to work: ...finished stopStone gemstone@ubuntu:~/GsDevKit_home$ stopNetldi seaside_ldi ================= GsDevKit script: stopNetldi seaside_ldi path: /home/gemstone/GsDevKit_home/bin/stopNetldi ================= stopping netldi seaside_ldi $GEMSTONE environment not set correctly It looks like I missed a step somewhere to set the $GEMSTONE var. what should it be set to? On Wed, Oct 21, 2015 at 1:07 PM, Dale Henrichs <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
On 10/21/2015 02:16 PM, Jon Paynter wrote: > Ok. so I went down the wrong road here... > > If putting everything under GsDevKit_home is preferred, then i will > use that (as I stated above, there is no old data that needs saving). > I was under the (false) impression /op/gemstone and GsDevKit shared files. > > So to avoid future confusion, I removed /opt/gemstone from my VM, and > re-ran the installServer from a fresh clone. I also had to update my > profile to remove references to /opt/gemstone > > The stone seems to be running according to the 'stones' command. > > next I need to start netldi with the correct port, so I stopped the > stone, but I cannot get the stopNetldi to work: > ...finished stopStone > gemstone@ubuntu:~/GsDevKit_home$ stopNetldi seaside_ldi knows the name of the netldi that it is associated with (via the TDSession). You can use the -b option on the startStone and stopStone commands to stop both the stone and the netldi ... > ================= > GsDevKit script: stopNetldi seaside_ldi > path: /home/gemstone/GsDevKit_home/bin/stopNetldi > ================= > stopping netldi seaside_ldi > $GEMSTONE environment not set correctly > > > It looks like I missed a step somewhere to set the $GEMSTONE var. > what should it be set to? may have helped you recognize ... I do recommend that you use the -h option to review the commandline arguments and options for the various commands ... I know that the GemStone stopnetldi command takes the name of the netldi and can lead to confusion when you "know too much:)" Dale _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
yeah it seems knowing too much is getting in the way. I keep wanting to start netldi and then stone, and stop them in reverse order. but in order to use ssh port forwarding for tode, you will need a way to specify the netldi port number on the startStone cmd, so for now I have to use this:./shared/downloads/products/GemStone64Bit3.2.8-x86_64.Linux/bin/startnetldi -a gemstone -g -P 55278 seaside_ldi project entry --git=builderGame --repo=$GS_HOME /home/external Topez ERROR: a TodeCommandError occurred (error 121000), Unknown option: git On Wed, Oct 21, 2015 at 3:08 PM, Dale Henrichs <[hidden email]> wrote:
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
|
|
On 10/22/2015 02:08 PM, Jon Paynter wrote: > yeah it seems knowing too much is getting in the way. I keep wanting > to start netldi and then stone, and stop them in reverse order. > > but in order to use ssh port forwarding for tode, you will need a way > to specify the netldi port number on the startStone cmd, so for now I > have to use this: > ./shared/downloads/products/GemStone64Bit3.2.8-x86_64.Linux/bin/startnetldi > -a gemstone -g -P 55278 seaside_ldi Yes, and the GsDevKit_home way is to set the netLDIPort field in the session description to 55278 and then the `startNetldi <stone-name>` command will look up the port number in the session description. > > So - on the VM here at work, the tode project list is functional and > im trying to create a new/blank project and add stuff. According to > the "project entry" man page, II should be able to run this: > project entry --git=builderGame --repo=$GS_HOME > /home/externalAnd create a new project. but when I put that in the > tode shell, i get an error: > Topez ERROR: a TodeCommandError occurred (error 121000), Unknown > option: git > > Is the git option not finished? or is it that it cant find git on the > host? bit as I've been distracted with other things both yesterday and today ... With that said this particular form of the `project entry` command is not for managing Smalltalk code. It is aimed at allowing you to manage a git repository that is not just used for code (like GsDevKit_home ... there is some Smalltalk code stored in the git repository for GsDevKit_home, but the main content of GsDevKit_home is shell and tODE scripts) ... and in your project list you'll see an example called `gsDevKit_home` that is attached to the GsDevKit_home git repo ... you can't load code from these guys, but you can `save` and `pull`, etc. > I know a few of the basic git commands, but virtually nothing about > MetaCello and how to use it. Hmmm okay ... There is a chapter on Metacello[1] in the Pharo By Example site that should give you some background on Metacello ... it does not cover a lot of the newer features of Metacello ... there is a section on creating a Baseline here[2] and there is a "Metacello User Guide"[3] that covers a lot of the newer features of Metacello. The there's the "Lock Command Reference"[4] that has a number of a bit more detailed examples that you can learn from ... Right now tODE is a bit sparse for absolute beginners, but I will be writing a `project new` command plus some documentation REAL SOON NOW, as I am on the final approach to release and there are a handful of things on my todo list that I must finish before the end of the month:) [1] http://pharobyexample.org/drafts/Metacello.pdf [2] https://github.com/dalehenrich/metacello-work/blob/master/docs/GettingStartedWithGitHub.md#create-baseline [3] https://github.com/dalehenrich/metacello-work/blob/master/docs/MetacelloUserGuide.md#metacello-user-guide [4] https://github.com/dalehenrich/metacello-work/blob/master/docs/LockCommandReference.md#lock-command-reference > > Or would it be better to create a blank project by creating stuff in > the file system somehow? > Okay ... The best way to get started modulo a bunch of study is to download and install the Sample project. Start by cloning the sample project: cd $GS_HOME/shared/repos git clone https://github.com/dalehenrich/sample.git Then create a project entry for the sample project in the tODE shell: project entry --baseline=Sample --repo=filetree://$GS_HOME/shared/repos/sample/repository --loads=`#('default')` /sys/stone/projects project list You should see the sample project in the project list (underlined because it is not loaded). Then load the Sample project: project load Sample project list Now you should see the Sample project loaded in the project list (bold because it is active). Use the "Browse > baseline" menu to bring up the class browse on the BaselineOfSample project ... in the baseline: method you see a specification for two packages: Sample-Core and Sample-Tests ... Sample-Core will be loaded before Sample-Tests, but for now you can ignore Sample-Tests (it wasn't loaded by default ... only Sample-Core was actually loaded). Use the "Browse > * " menu items to look at the Sample-Core category and package ... go ahead and edit the SampleCore class and refresh the project list to see that the dirty projects are marked with a * ... go ahead and " save " the project to see what happens and you can visit the git repository in the bash shell to look at the disk layout ... Ask questions about the baseline and anything else that you might have ... you could even start working on your project and just use the sample project for storing checkpoints on disk ... in a follow-on message I'll describe the steps for creating a BaselineOf of your very own, a package and git repository of your very own (and we can move your classes to the new packages ... but this should keep you busy for a little while:) Maybe I'll just whip up the `project new command this afternoon/eveining:) Dale _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
«
Return to GLASS
|
1 view|%1 views
| Free forum by Nabble | Edit this page |