Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
Hi list, At first, i setup a 3 nodes cluster and use the default lvm storage plugin and has a test resource/volume works. However, when I read the document I found that lvm storage plugin doesn't support drbdmanage snapshots, so I'm thinking to change storage plugin to ThinLV on all 3 nodes. I did these below steps and finally I got error like below. ``` [root at mesos-test-dev001-whdx ~]# drbdmanage list-assignments -p +------------------------------------------------------------------------------------------------------+ | Node | Resource | Vol ID | State | | ------------------------------------------------------------------------------------------------------ | | mesos-test-dev001-whdx.qiyi.virtual | r0 | * | FAILED(3), pending actions: commission | | mesos-test-dev001-whdx.qiyi.virtual | r0 | 0 | pending actions: commission, attach | | mesos-test-dev002-whdx.qiyi.virtual | r0 | * | FAILED(3), pending actions: commission | | mesos-test-dev002-whdx.qiyi.virtual | r0 | 0 | pending actions: commission, attach | | mesos-training-dev001-cqdx.qiyi.virtual | r0 | * | FAILED(3), pending actions: commission | | mesos-training-dev001-cqdx.qiyi.virtual | r0 | 0 | pending actions: commission, attach | +------------------------------------------------------------------------------------------------------+ ``` So I'm wondering if it is possible to change nodes storage plugin on the fly? without re-initialize the whole cluster. Details like below. 1. delete all resources and volumes use drbdmanage remove-volume and remove-resource 2. use drbdmanage modify-config to change every nodes storage plugin to ThinLV like below ``` [Node:mesos-training-dev001-cqdx.qiyi.virtual] storage-plugin = drbdmanage.storage.lvm.ThinLV [Node:mesos-test-dev002-whdx.qiyi.virtual] storage-plugin = drbdmanage.storage.lvm.ThinLV [Node:mesos-test-dev001-whdx.qiyi.virtual] storage-plugin = drbdmanage.storage.lvm.ThinLV ``` 3. create resource and volume and deploy to 3 nodes and I got the above error and after that, I did try `drbdmanage reconfigure and restart`, but seems nothing help. -- thanks, chengwei