mirrors/ceph-csi
1
0
Fork 0
mirror of https://github.com/ceph/ceph-csi.git synced 2024-12-28 16:00:22 +00:00
Code Issues Packages Projects Releases Wiki Activity
289 Commits 31 Branches 40 Tags 155 MiB
1a246f7bf3
Commit Graph

124 Commits

Author SHA1 Message Date
Madhu Rajanna
1a246f7bf3 simplify the error check 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:16:03 +05:30
Madhu Rajanna
36f99e36ca Fix unparam issues 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:16:03 +05:30
Madhu Rajanna
284c5801c3 Fix golint issue 
pkg/rbd/rbd.go:67:65⚠️ exported func NewNodeServer
returns unexported type *rbd.nodeServer, which can be
annoying to use (golint)

Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:16:03 +05:30
Madhu Rajanna
4c564207c6 Fix error checking issue in json Decode 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:15:25 +05:30
Madhu Rajanna
1d11d0acc3 Fix lint issues 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:15:25 +05:30
Madhu Rajanna
5eb1974e38 Fix vetshadow issues 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:15:25 +05:30
Madhu Rajanna
75beccd8c0 Remove unused variable 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:14:48 +05:30
Madhu Rajanna
49c8fe5d50 Remove conversion which is not required 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:14:48 +05:30
Madhu Rajanna
1c03334a90 Add constant for rbd and ndb 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:14:48 +05:30
Madhu Rajanna
fa78c00535 update struct names and struct fiels 
it wont be meaningful to call cephfs.NewcephfsDriver()
to get a new driver, it will be better if we call
cephfs.GetNewDriver() which returns the cephfs driver
object.

same goes for rbd also

Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:14:48 +05:30
Madhu Rajanna
15b5b0112e rename Id to ID to fix lint issue 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:14:48 +05:30
Madhu Rajanna
8f2941dd47 Remove code complexity 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:13:04 +05:30
Madhu Rajanna
de1f099680 Remove unused struct fields 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:10:52 +05:30
Madhu Rajanna
9f76f6bd59 Remove dead code 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-25 14:10:52 +05:30
Huamin Chen
c6c496ff59 switch to node registrar 2019-01-22 14:46:41 -05:00
Huamin Chen
0151792684 review feedback: make monValueFromSecret override monitors if both are set 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-21 09:21:03 -05:00
Huamin Chen
db463edeef allow ceph mon stored in secret so when mon changes, cephfs driver can get latest mons and override old ones 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-18 10:27:48 -05:00
Huamin Chen
48407e2484 add csi volume device mount path to csi plugin 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-17 08:57:18 -05:00
Huamin Chen
263c45bb45 enable csi block; use canary external-provisioner image to pick up block volume provisioning 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-16 13:52:45 -05:00
Masaki Kimura
0e60dabca3 Move resolving bind mount logic from k8s 2019-01-16 13:00:48 -05:00
Masaki Kimura
ea75a9d162 Fix pv deletion issue caused by not dettaching due to wrong mount count 2019-01-16 12:51:13 -05:00
Masaki Kimura
5867d495fd Change csi rbd to create/delete targetPath for publish/unpublish 2019-01-16 12:50:17 -05:00
Masaki Kimura
165b82a44c Add block supports to rbd driver 2019-01-16 12:49:02 -05:00
Madhu Rajanna
e42e66ff30 user errors.New if error formatting is not required 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-16 18:54:13 +05:30
Madhu Rajanna
17f5c0a7ce use Errorf is error formatting is required 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-16 18:54:06 +05:30
Madhu Rajanna
20af5afcab Fix golint issues 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-16 18:33:38 +05:30
Madhu Rajanna
9ddc265c10 reject block volume creation in cephfs 
Signed-off-by: Madhu Rajanna <mrajanna@redhat.com>
 2019-01-16 18:17:14 +05:30
Huamin Chen
aed7506d88 fix merge leftovers; use canary driver-registrar image, as v1.0.0 is not hosted in quay.io 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-15 13:31:06 -05:00
Huamin Chen
e46099a504 reconcile merge 
Signed-off-by: Huamin Chen <hchen@redhat.com>
 2019-01-15 16:20:41 +00:00
Huamin Chen
85b8415024 Merge branch 'master' into master-to-1.0 2019-01-15 16:15:30 +00:00
mickymiek
b387daaabf remove useless comment 2019-01-14 20:15:09 +00:00
mickymiek
32cb974b8c gofmt 2019-01-14 20:15:09 +00:00
mickymiek
62d65ad0cb cm metadata persist for rbd and cephfs 2019-01-14 20:15:09 +00:00
Patrick Ohly
51d6ac6f55 rbd: refuse to create block volumes 
Without this check, the driver fails one of the E2E storage tests in
Kubernetes 1.13: provisioning a block volume is expected to fail in
e689d515f7/test/e2e/storage/testsuites/volumemode.go (L329-L330)
 2019-01-14 20:15:09 +00:00
gman
29bdeb2261 cephfs: don't set quotas for zero-sized volumes 2019-01-14 20:15:09 +00:00
Patrick Ohly
403cad682c 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.
 2019-01-14 20:15:09 +00:00
Mike Cronce
2c3961b960 pkg/rbd/rbd.go: Fix "go vet" errors 2018-12-04 21:54:00 -05:00
Mike Cronce
23a4126aed pkg/rbd/controllerserver.go: gofmt 2018-12-04 21:44:04 -05:00
Mike Cronce
d9fbdeb517 pkg/cephfs: Use request name to generate deterministic volume names 2018-12-04 21:39:00 -05:00
Mike Cronce
04872e5ebf Merge branch 'master' of github.com:ceph/ceph-csi into csi-v1.0.0 2018-12-04 16:28:37 -05:00
Mike Cronce
22e23640a4 pkg/rbd/rbd.go: Updated PluginFolder to use new CSI 1.x directory 2018-12-04 15:38:09 -05:00
Mike Cronce
37caeb5b2c pkg/cephfs/driver.go: Updated PluginFolder to use new CSI 1.x directory 2018-12-04 15:38:09 -05:00
gman
ed811e0506 cephfs: don't set quotas for zero-sized volumes 2018-12-01 10:39:09 +01:00
Mike Cronce
af3083f717 pkg: Updated "version" variables from 0.3.0 to 1.0.0 2018-11-29 13:15:52 -05:00
Mike Cronce
93cb8a04d7 pkg/rbd: Updated for new versions of CSI/Kubernetes dependencies 2018-11-24 14:18:24 -05:00
Mike Cronce
41b30eb6c2 pkg/cephfs: Updated for new versions of CSI/Kubernetes dependencies 2018-11-24 13:48:36 -05:00
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