Jeff Mitchell
fc6564e4ee
Don't run rollback and upgrade functionality if we are a replication secondary ( #3900 )
...
* Don't run rollback and upgrade functionality if we are a replication
secondary, but do if the mount is local.
2018-02-02 20:28:25 -05:00
Brian Kassouf
2f19de0305
Add context to storage backends and wire it through a lot of places ( #3817 )
2018-01-19 01:44:44 -05:00
Brian Kassouf
1c190d4bda
Pass context to backends ( #3750 )
...
* Start work on passing context to backends
* More work on passing context
* Unindent logical system
* Unindent token store
* Unindent passthrough
* Unindent cubbyhole
* Fix tests
* use requestContext in rollback and expiration managers
2018-01-08 10:31:38 -08:00
Chris Hoffman
6651f3aa33
checking for nil backends before attempting to rollback on a backend ( #2997 )
2017-07-12 13:12:55 -07:00
Jeff Mitchell
8091a10c38
Make rollback attempts trace level instead of debug level
2017-02-27 09:41:56 -05:00
Jeff Mitchell
fffee5611a
Rejig locks during unmount/remount. ( #1855 )
2016-09-13 11:50:14 -04:00
Jeff Mitchell
58b32e5432
Convert to logxi
2016-08-21 18:13:37 -04:00
Jeff Mitchell
80a688c059
Ensure mount/auth tables are not nil when triggering rollback
...
During setup or teardown there could be a race condition so check for it
to avoid a potential panic.
2016-07-18 22:02:39 -04:00
vishalnayak
1d94828e45
Re-enable rollback triggers for auth backends
2016-05-26 14:29:41 -04:00
Jeff Mitchell
417a56c42b
Disable rollback on auth for now and add workaround for its auth/ adding to entry paths
2016-05-25 17:53:45 -04:00
vishalnayak
9aa8fb6cc1
Support periodic tidy callback and config endpoints.
2016-04-26 10:22:29 -04:00
Jeff Mitchell
bece637eb7
Address feedback from review
2015-11-15 17:32:57 -05:00
Jeff Mitchell
bc4c18a1cf
Rearchitect MountTable locking and fix rollback.
...
The rollback manager was using a saved MountTable rather than the
current table, causing it to attempt to rollback unmounted mounts, and
never rollback new mounts.
In fixing this, it became clear that bad things could happen to the
mount table...the table itself could be locked, but the table pointer
(which is what the rollback manager needs) could be modified at any time
without locking. This commit therefore also returns locking to a mutex
outside the table instead of inside, and plumbs RLock/RUnlock through to
the various places that are reading the table but not holding a write
lock.
Both unit tests and race detection pass.
Fixes #771
2015-11-11 11:54:52 -05:00
Armon Dadgar
496ebe561c
vault: cleanups for the audit log changes
2015-06-29 15:27:28 -07:00
Nate Brown
c55f103c58
Adding error and remote_address to audit log lines
2015-06-18 17:17:18 -07:00
Armon Dadgar
512b3d7afd
vault: Adding metrics profiling
2015-04-08 16:43:17 -07:00
Armon Dadgar
d74c4c1c33
vault: Remove log about rollback
2015-04-03 17:11:24 -07:00
Armon Dadgar
f231a6c67d
vault: rollback supports joining an inflight operation
2015-04-01 22:12:03 -07:00
Mitchell Hashimoto
6f9d63dea5
vault: comment mounts mapping in rollback manager
2015-03-17 20:53:28 -05:00
Mitchell Hashimoto
05f86ca957
vault: put uint32 at top of struct to avoid alignment issues
2015-03-17 20:46:10 -05:00
Mitchell Hashimoto
97dab0c285
vault: ignore backends that don't support rollback
2015-03-17 20:39:45 -05:00
Mitchell Hashimoto
e078b957d4
vault: start/stop rollback manager post/pre seal
2015-03-17 20:39:45 -05:00
Mitchell Hashimoto
c7b9148841
vault: RollbackManager
...
There are some major TODO items here, and it isn't hooked into the core
yet, but the basic functionality is there.
2015-03-17 20:39:45 -05:00
