Cannot update workspace "sql: Scan error on column index 0: converting driver"

Pydio Cells
1

Hi,

I cannot update or create new workspace. I have an SQL Error

I use an Percona XtraDB Cluster with 3 servers and a distributed Minio with 4 servers all in docker swarms 18.03.

I cannot find the “cnil” string in the sql DUMP or in the name of the files.

No error in Percona logs neither in Minio logs.

|PYDIO_cells.1.44p0e09tw5ui@scaleway3 | 2018-08-15T10:38:16.849Z|INFO|pydio.grpc.tasks|Run Job internal-prune-jobs on timer event Iso8601Schedule:“R/2012-06-04T19:25:16.828696-07:03/PT10M” |
|—|---|—|---|
|PYDIO_cells.1.44p0e09tw5ui@scaleway3 | 2018-08-15T10:38:16.850Z|INFO|pydio.grpc.tasks|Run Job archive-changes-job on timer event Iso8601Schedule:“R/2012-06-04T19:25:16.828696-07:03/PT10M” |
|PYDIO_cells.1.44p0e09tw5ui@scaleway3 | 2018-08-15T10:38:17.895Z|ERROR|pydio.grpc.tasks|{“id”:"",“code”:0,“detail”:“sql: Scan error on column index 0: converting driver.Value type \u003cnil\u003e (”\u003cnil\u003e") to a uint64: invalid syntax","|

Best regards

PS : french user

2

Hello,

I would rather say that \u003cnil\u003e is the encoded version of <nil>. Could you please give a little bit more info about the configuration of Cells that you use, so that we might help to fix this driver issue?

Thanks.

3

Hi,

I’m on holidays.

Here are the configuration.

4 servers C2M from Scaleway

Docker version 18.03 since there is an issue with minio and Docker 10.06 reported by here https://github.com/minio/minio/issues/6240

One Swarm with 4 nodes

First stack ;

4 nodes of Minio : one node on each server RELEASE.2018-08-02T23-11-36Z

Second stack :

3 nodes of etcd

3 nodes of Percona XtraDB Cluster with enforcing mode and disable strict SQL for running Pydio. Current release

Third stack :

1 node with pydio cell image

1 node with your phpfpm image

The files were alreay copied into a bucket on the Minio.

I create a Datasource and a workspace.

I launch the sync.

Best regards.

Sébastien

4

Hi Sebastien,

Just for you to know, I’ve fixed a few issues today for next patch release and among others the one that had thrown the error you reported.
It was an initialisation issue for a scheduler task that archive change events. This happens when the base is empty.

That means that we will have to investigate in another direction to solve your issue when you are back.
Until there, I hope your enjoy your holidays!

Cheers.