Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • 12:36

    plyhun on hotfix-1.4.x-pdf-metadata

    Fix uploading of PDFs with comp… (compare)

  • 09:48

    Jotschi on dev-plugin-local-client

    (compare)

  • 09:48
    Jotschi closed #941
  • 09:48

    Jotschi on dev-reduce-memory

    (compare)

  • 09:48
    Jotschi closed #893
  • 09:47

    Jotschi on hotfix-1.4.x-thread-blocked-during-startup

    (compare)

  • 09:47
    Jotschi closed #1036
  • 09:47
    Jotschi closed #962
  • 09:46

    Jotschi on dev-docker-clustering

    (compare)

  • 09:46
    Jotschi closed #410
  • 09:46

    Jotschi on dev-jmh

    (compare)

  • 09:46

    Jotschi on micronode-raw-field

    (compare)

  • 09:46
    Jotschi closed #576
  • 09:46
    Jotschi closed #474
  • 09:45

    Jotschi on dev-madl

    (compare)

  • 09:45
    Jotschi closed #634
  • 09:45

    Jotschi on dev-orientdb-3.0.0

    (compare)

  • 09:45
    Jotschi closed #436
  • 09:45
    Jotschi closed #896
  • 09:45
    Jotschi closed #956
Johannes Schüth
@Jotschi
If you use the plugin api you can access the eventbus via java and also subscribe only to "local" events. Via REST / websocket this is however not possible.
Johannes Schüth
@Jotschi
Instance B would not emit the event. In this case Instance A would emit the event but pass the event along to any subscriber
Most events also contain an "origin" in the body. This can be used to identify from which instance the received event originated.
René Calles
@ReneVolution
Thanks @Jotschi, that was still a bit confusing to me, so I basically connected to each instance individually and could confirm that an event is only emitted once in the eventbus of the instance which received the request to e.g. create or update a node. All other would be silent. Therefore, in my use case, I will need to have individual eventbus clients connect to their assigned instance. Otherwise I would miss messages I am listening for. Thanks a lot for taking the time explaining.
Johannes Schüth
@Jotschi
@ReneVolution It also depends on what method you are using to dispatch a message. Messages via "send" will only be received by one receiver. And messages which have been send via publish will be received by all clients that subscribed to the specific address. I'm not sure how this translates on the websocket but thats the basic rule for the eventbus in general.
tgnstr
@tillganster
Hello together, I have a REST-Plugin which uses the admin client to create nodes and set permissions on this nodes, and this worked for over a year. But suddenly the createNode(..) request of the admin client gets a 401 error. How is this even possible ? Is this a known problem or has someone an idea where to search for this failure ?
Johannes Schüth
@Jotschi
I'm not aware of such a problem.
micstra
@micstra
hello, i'm currently facing problems when migrating from version 1.7.1 to 1.7.2 concerning this github issue gentics/mesh#1152. is there any workaround to still do the migration? export/import does not really work, since the server tries to apply the migration on the first restart and then crashes again.
Johannes Schüth
@Jotschi
@micstra Have you tried to update from 1.7.1 to 1.7.4 ? I did not yet find time to look into this issue.
micstra
@micstra
yes, unfortunately the same error
Johannes Schüth
@Jotschi
I may remove the changelog entry entirely and replace it with a on-the-fly migration of the database.
But I'm not sure when I'll be able to work on this.
micstra
@micstra
ok thanks for the fast response :) is there a possibility to start the high-level-migration through the api? my last idea would be running an empty 1.7.2 instance, start the import and then manually trigger the migration. this would probably resolve the issue with the non-active/creatable transaction
Johannes Schüth
@Jotschi
No unfortunately there is no api endpoint to manage the changelog. I also had the idea to add this in the past but I never managed to get around to add it.
DevNils
@DevNils
Hello Folks, I got a question regarding export/import. I am running mesh in docker. I went to machine A and triggered a graphdb/export which was successful. I copied the export json.gz and all upload data from the docker volumes. Then I started a brand new instance on another machine B (I also tried a used one) and copied everything to the correct directories in the volumes. I executed a graphdb/importwhich was successful. I checked the logs as well. It seems that now something is messed up with users and permissions. I am not able to login anymore to mesh on machine B. I tried all possible passwords from machine A and B. Do you have an idea what I did wrong? Are there any special parameters I have to pass on export/import?
DevNils
@DevNils
PS: I am using mesh 1.7.4
Johannes Schüth
@Jotschi
Did restart of the mesh instance solve the problem for you?
DevNils
@DevNils
Thanks a lot, that solved it!
DevNils
@DevNils
It is me again ;) Everything was fine until I tried to upload new binary data. I guess there is a problem with access rights of the copied uploads data. Is there something special I need to do when I use docker cp to copy data from one docker volume to another?
The Error is
io.vertx.core.file.FileSystemException: java.nio.file.AccessDeniedException: /uploads/temp/c503903c3d4146d68b9152ef26d65a54.tmp
    at io.vertx.core.file.impl.FileSystemImpl$3.perform(FileSystemImpl.java:500)
    at io.vertx.core.file.impl.FileSystemImpl$3.perform(FileSystemImpl.java:493)
    at io.vertx.core.file.impl.FileSystemImpl$BlockingAction.handle(FileSystemImpl.java:974)
    at io.vertx.core.file.impl.FileSystemImpl$BlockingAction.handle(FileSystemImpl.java:955)
    at io.vertx.core.impl.ContextImpl.lambda$executeBlocking$2(ContextImpl.java:316)
    at io.vertx.core.impl.TaskQueue.run(TaskQueue.java:76)
    at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
    at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
    at io.netty.util.concurrent.FastThreadLocalRunnable.run(FastThreadLocalRunnable.java:30)
    at java.base/java.lang.Thread.run(Unknown Source)
Caused by: java.nio.file.AccessDeniedException: /uploads/temp/c503903c3d4146d68b9152ef26d65a54.tmp
    at java.base/sun.nio.fs.UnixException.translateToIOException(Unknown Source)
    at java.base/sun.nio.fs.UnixException.rethrowAsIOException(Unknown Source)
    at java.base/sun.nio.fs.UnixException.rethrowAsIOException(Unknown Source)
    at java.base/sun.nio.fs.UnixCopyFile.copyFile(Unknown Source)
    at java.base/sun.nio.fs.UnixCopyFile.move(Unknown Source)
    at java.base/sun.nio.fs.UnixFileSystemProvider.move(Unknown Source)
    at java.base/java.nio.file.Files.move(Unknown Source)
    at io.vertx.core.file.impl.FileSystemImpl$3.perform(FileSystemImpl.java:498)
    ... 9 common frames omitted
DevNils
@DevNils
PS: Solved it by my self. When you do this you should not copy the temp directory from uploads!
Johannes Schüth
@Jotschi
@DevNils Strange. Maybe the permissions changed?
@micstra The next version of Gentics Mesh will remove the error prone changelog entry as previously mentioned. I'm currently finalizing the change.
Sorry that it took a while but I'm also busy with non-mesh related tasks.
Onno Schmidt
@ezonno
Hi is there some example available on how to search for text inside binary_content ?
Johannes Schüth
@Jotschi

@ezonno I think there is no full example for this in the docs but I have uploaded this pdf: http://www.africau.edu/images/default/sample.pdf into the demo server. This results in this node: https://demo.getmesh.io/api/v2/demo/nodes/c8fa6c8a64594642b83df0624196acbf

The query on the nodes search endpoint could for example be something like:

{
    "query": {
        "bool": {
            "must": [
                {
                    "match": {
                        "schema.name.raw": "binary_content"
                    }
                },
                {
                    "match": {
                        "fields.binary.file.content": "Simple PDF"
                    }
                }
            ]
        }
    }
}

The important part is the field name in the ES document which is fields.binary.file.content.

Please note that the example / node may be removed in the demo. The demo will be reset everyday.

Additional information on binary search can be found in our docs: https://getmesh.io/docs/elasticsearch/#_binary_search

micstra
@micstra
@Jotschi thanks for the info! looking forward to 1.7.5 ;)
Onno Schmidt
@ezonno
@Jotschi Thank you for this search example that helps a lot!
Johannes Schüth
@Jotschi
@/all Gentics Mesh 1.7.5 has been released: https://getmesh.io/docs/changelog/#v1.7.5
micstra
@micstra
hello, its me again. is it still possible to submit contributions to the old frontend? as fas as i figured out it is the beta-branch on the mesh-ui repository
Johannes Schüth
@Jotschi
@micstra Yes, that should still be possible. We are however unable to port changes from the old ui to the new one.
micstra
@micstra
yeah understandable. ok so if submit a bugfix-pr to the beta branch and it is accepted, then it will be included in some future-release of mesh
Johannes Schüth
@Jotschi
Yes
micstra
@micstra
ok thanks :)
Johannes Schüth
@Jotschi
You can however also build the ui yourself and deploy it along with mesh in a dedicated container.
micstra
@micstra
i already thought about that however i actually dont want to stick with the old frontend forever. in the old-frontend the pagination is always reset to the first page, if you click on a node. i did not look into the code yet, but imo this is a not-so-difficult fix. thus i thought about contributing
Johannes Schüth
@Jotschi
I see. Maybe that's an easy fix. I don't know. I think the new UI may also had the issue. I recall that we did some fixes on the handling of paging. Maybe it may be useful to check the commits.
micstra
@micstra

i now figured out the bug i mentioned is already fixed. it was fixed with version 0.9.3

1b6df808 (HEAD -> beta, tag: 0.9.4, origin/beta) [Jenkins | beta] Release version 0.9.4
ce0447e6 Prepare release
9ebc52a6 Validate microschema lists, Node update error handling (#202)
4bf3ef84 (tag: 0.9.3) [Jenkins | beta] Release version 0.9.3
977ea3d3 Update version
e8e46bc3 Copy query params on schema search requests
eaa2153e Fix project explorer select node resetting current page; #194 (#201)
63111562 (tag: 0.9.2) [Jenkins | beta] Release version 0.9.2

the problem is that the backend (mesh-project) currently pins version 0.9.2 as dependency for the old frontend.

<mesh.admin-ui.version>0.9.2</mesh.admin-ui.version>
is there a particular reason why the dependency to the old-ui was reverted back to 0.9.2?
Johannes Schüth
@Jotschi
I'm not aware that it was reverted. I assume it may not have been updated. I'm not sure.
micstra
@micstra

hello again, the git diff tells me it was reverted

-               <mesh.admin-ui.version>0.9.4</mesh.admin-ui.version>
+               <projectRoot>${project.basedir}</projectRoot>
+               <mesh.admin-ui.version>0.9.2</mesh.admin-ui.version>
+               <mesh.admin-ui2.version>0.10.6</mesh.admin-ui2.version>

unfortunately the commit is over one year old so probably nobody remembers if this change was intentionally.
the commit hash is 8895f6817e with the commit message "WIP: Add development version of new UI to Mesh Server (#621)" hopefully this helps.

Johannes Schüth
@Jotschi
Strange - I'll ask around. Maybe somebody recalls why it was reverted.
Johannes Schüth
@Jotschi
@micstra Nobody recalls. We switched our focus on the new UI.
micstra
@micstra
thanks for asking. so i guess there is little chance that the pinned version of the old-ui will be upgraded?
Johannes Schüth
@Jotschi
I think the previous version introduced a regression via a contribution which we would need to be rolledback first. As previously stated we want to focus on the new ui.
micstra
@micstra
understandable. anyways thanks for the help and responsiveness!
Florian
@flobauer
Hello, I have some difficulties to understand Permissions. When I want to give somebody only the permission to do whatever he/she wants in a subfolder of a project and only on that particular folder I use MeshUI to add the permissions to the roles to the group and add the users to it.
When I give edit/delete rights to the specific folder/node and apply it recursively the user can navigate to the folder but cannot edit any contents. Is there a documentation or example that can help me guide through that?
Florian
@flobauer
image.png
The permissions I get for that specific user look good:
Florian
@flobauer
Ok, I got it the work -> read access to the schemas seems to be needed