Commit Graph

125 Commits

Author SHA1 Message Date
Patrick Ohly
720ad4afeb rbd: protect against concurrent gRPC calls
The timeout value in external-provisioner is fairly low. It's not
uncommon that it times out and retries before the rbdplugin is done
with CreateVolume. rbdplugin has to serialize calls and ensure that
they are idempotent to deal with this.
2018-10-26 15:29:48 +02:00
Huamin Chen
188cdd1d68
Merge pull request #89 from rootfs/containerized
support nsmounter when running in containerized mode
2018-10-15 20:25:40 -04:00
Huamin Chen
3436a094f7 support nsmounter when running in containerized mode
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-10-15 14:59:41 +00:00
Patrick Ohly
25e3a961c3 rbdplugin: idempotent DeleteVolume
When the initial DeleteVolume times out (as it does on slow clusters
due to the low 10 second limit), the external-provisioner calls it
again. The CSI standard requires the second call to succeed if the
volume has been deleted in the meantime. This didn't work because
DeleteVolume returned an error when failing to find the volume info
file:

  rbdplugin: E1008 08:05:35.631783       1 utils.go:100] GRPC error: rbd: open err /var/lib/kubelet/plugins/csi-rbdplugin/controller/csi-rbd-622a252c-cad0-11e8-9112-deadbeef0101.json/open /var/lib/kubelet/plugins/csi-rbdplugin/controller/csi-rbd-622a252c-cad0-11e8-9112-deadbeef0101.json: no such file or directory

The fix is to treat a missing volume info file as "volume already
deleted" and return success. To detect this, the original os error
must be wrapped, otherwise the caller of loadVolInfo cannot determine
the root cause.

Note that further work may be needed to make the driver really
resilient, for example there are probably concurrency issues.
But for now this fixes: #82
2018-10-09 12:08:56 +02:00
Huamin Chen
239f295dd1
Merge pull request #79 from rootfs/rbd-nbd
allow monitors be embedded in credential secret
2018-09-24 08:58:09 -04:00
Huamin Chen
d5b7543565 allow monitors be embedded in credential secret
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-09-21 14:43:01 +00:00
Kenjiro Nakayama
c1e072de0b Fix misspelling of "successfully" 2018-09-21 23:08:23 +09:00
Huamin Chen
30a5d9a6e7 add rbd-nbd mounter in storage class
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-09-18 14:09:12 +00:00
Huamin Chen
6f3625b11e review feedback
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-09-18 13:10:28 +00:00
Huamin Chen
8955eb03bc support rbd-nbd
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-09-17 18:12:22 +00:00
Seungcheol Ko
38aa575925 check snapshot feature 2018-08-09 22:07:13 +09:00
Seungcheol Ko
4312907f7b remove the snapshot if can't store snapshot information 2018-08-09 22:07:06 +09:00
Seungcheol Ko
b0e68a52e0 Refactoring using users 2018-08-09 22:07:00 +09:00
Seungcheol Ko
7d90783f03 fix nit 2018-08-09 22:06:51 +09:00
Seungcheol Ko
f0fba1240a Revert "Implement NodeGetInfo for csi spec 3.0"
This reverts commit c93466b009.
2018-08-08 20:22:59 +09:00
Seungcheol Ko
b1ccdbb154 Support snapshot feature in rbdplugin 2018-08-08 17:16:07 +09:00
Seungcheol Ko
c93466b009 Implement NodeGetInfo for csi spec 3.0 2018-08-08 14:41:45 +09:00
Seungcheol Ko
bc34bd389e support image features for csi-rbdplugin 2018-07-21 00:59:54 +09:00
malc0lm
f273874f26 rbd: advertises PluginCapability_Service_CONTROLLER_SERVICE 2018-06-13 15:14:15 +08:00
Serguei Bezverkhi
1ff1f8be30 code update 2018-03-08 13:22:40 -05:00
Huamin Chen
9d4b49b54b node server: don't persist vol
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-02-20 16:10:59 +00:00
Serguei Bezverkhi
de16b0f5e9 Refactor rbdplugin for CSI 0.2.0 spec 2018-02-15 10:49:38 -05:00
Huamin Chen
57a4715e95 rbd plugin refactoring: remove k8s clientSet
Signed-off-by: Huamin Chen <hchen@redhat.com>
2018-02-06 14:36:58 +00:00
Serguei Bezverkhi
8614aee8ef Making VolumeID != volName as per CSI spec 2018-01-15 20:57:07 -05:00
Serguei Bezverkhi
27334ac60d just code files 2018-01-15 11:50:50 -05:00