xCAT/confluent
2
0
Fork 0
mirror of https://github.com/xcat2/confluent.git synced 2025-04-18 11:09:30 +00:00
Code Issues Projects Releases Wiki Activity
1,606 Commits 39 Branches 67 Tags
Commit Graph

124 Commits

Author SHA1 Message Date
Jarrod Johnson
6aec9534e7 Fixes for nodesupport 2018-08-23 16:56:40 -04:00
Jarrod Johnson
3ee6334db2 Fix transfer owner argument for servicedata 2018-08-23 16:48:10 -04:00
Jarrod Johnson
582a4de62d Add CLI and directory support for nodesupport 2018-08-23 16:36:41 -04:00
Jarrod Johnson
c9959d4082 More of the service data retrieval api 
More progress is made toward the goal
 2018-08-23 16:14:49 -04:00
Jarrod Johnson
fa11fb54cb Add API support for getting service data 
Service data retrieval is a common activity required
for interacting with support.
 2018-08-23 15:39:25 -04:00
Jarrod Johnson
ee3b824870 Add /description to nodes api 2018-08-23 08:44:41 -04:00
Jarrod Johnson
8787d23b3a Add advanced to API for system configuration. 
pyghmi makes hiding optional advanced settings.
 2018-08-17 10:59:50 -04:00
Jarrod Johnson
9b48110155 Do not proceed a logged, but broken session 
It shouldn't be possible for this to be the case, but out of an
abundance of caution, check for this.  So far only produced this by
forcing broken = True in a debug session.  Intended to catch an alleged
scenario where console was managing to use a broken session (fixed in
pyghmi) and have confluent also recognize the situation for non-console
usage).
 2018-08-16 14:43:16 -04:00
Jarrod Johnson
1d4df8af3a Fix extraneous error in log on connectivity loss 2018-08-07 15:43:53 -04:00
Jarrod Johnson
2aba6e469c Correct variable name in the 'connected' fix 2018-08-07 15:31:41 -04:00
Jarrod Johnson
de58593f14 Fix inability to notice underlying broken layers of the SOL 
Through an unknown set of circumstances, an solconnection could be
stuck 'connecting'.  In every case analyzed, the ipmi_session was
broken.  Use that to detect a class of failure and react appropriately.
 2018-08-07 15:12:53 -04:00
Jarrod Johnson
3ff7e42074 Change behavior for fallback handling 
Fallback would do nothing to fix a persistent problem with an IPMI
session.  For lack of knowing how to avoid the situation, at least
make changes so it won't go wrong in the future.
 2018-07-20 13:20:50 -04:00
Jarrod Johnson
fab177e077 Fix node[group][attrib|define] handling of = 
Attributes with = in the value were not handled correctly,
fix by only doing one split.
 2018-07-20 09:54:17 -04:00
Jarrod Johnson
0b26d12837 Fix memory leaks 
For one, configmanager was left with stale callback references, clean
those up.

For another, the callback pattern was creating a circular reference that
python memory management couldn't overcome.  Break the reference
explicity when an item is disposed of.
 2018-07-03 08:58:11 -04:00
Jarrod Johnson
003358da5f Discontinue ipmi session on collective manager change 2018-06-19 15:10:59 -04:00
Jarrod Johnson
675dc966c7 Clean up errors on bad ipv4 addresses 
confluent errors are better now
 2018-05-17 16:24:06 -04:00
Jarrod Johnson
66e8ce2dde Merge branch 'master' of github.com:jjohnson42/confluent 2018-03-27 16:35:31 -04:00
Jarrod Johnson
f97c39cea4 Add hostname to api 
The hostname of the BMC is added to the api.
 2018-03-27 15:51:14 -04:00
Jarrod Johnson
6671b9aad3 Provide cleaner behavior on timeouts 
If a timeout occurred outside of a keeplaive, provide
a more consistent message about the situation.
 2018-03-23 08:27:27 -04:00
Jarrod Johnson
9269c9feff Provide more reasonable error message to user 
If a protocol is unsupported, propogate error message usefully.
 2018-02-27 14:51:00 -05:00
Jarrod Johnson
0fd2d26f82 Add access to nodefirmware shortcut 
The nodefirmware can now request firmware more selectively.
 2018-02-14 14:38:20 -05:00
Jarrod Johnson
453f6d8016 Catch general SSL errors to BMC 
Provide better info on general SSL misbehaviors.
 2018-02-13 13:17:29 -05:00
Jarrod Johnson
685f5c6803 Add system configuration 
This gives access to system configuration where available
 2018-02-02 16:20:13 -05:00
Jarrod Johnson
7f3763f9eb Clean up media error handling 2018-01-30 15:18:20 -05:00
Jarrod Johnson
ba0e03d454 Provide a catchall for general certificate mismatch 
Beyond the special partial cases for firmware/hardware inventory,
provide a catchall for invalid public keys and a message for it.
 2018-01-29 14:17:01 -05:00
Jarrod Johnson
e193e92fbf Make TLS errors in hardware/firmware more consistent 
Also, place the error near the end where it is more likely to be
noticed, and close to where user would have expected the data to be.
 2018-01-29 14:12:34 -05:00
Jarrod Johnson
a7b0a449d6 Put ipmi errors in trace log 
Put an end to the 'see stderr log'
 2018-01-29 13:21:18 -05:00
Jarrod Johnson
0d40a0cac6 Add ability to clear upload progress tracking 
Remote media uploads could not be cleared, add the ability to do so.
 2017-12-08 18:58:04 -05:00
Jarrod Johnson
6845f64d46 Correct spelling of tenant attribute 
The attribute had been mispelled and needed a fix.
 2017-12-08 18:42:08 -05:00
Jarrod Johnson
09d20ea1ff Add remote media handling to API 2017-12-08 15:03:27 -05:00
Jarrod Johnson
2c5432454a Implement a nicer nodereseat unsupported command 
This carries the error condition cleanly back to the command line.
 2017-11-28 13:42:21 -05:00
Jarrod Johnson
52673a990b Confluent server side enclosure reseat 
This provides ability to request reseat of nodes, and redirects it to
the enclosure manager.
 2017-10-26 11:45:16 -04:00
Jarrod Johnson
3372a8401a Move registration of key handler up to constructor 
This should be a more bulletproof place to be.  Note that it used
to be here and was moved because pyghmi used to call oem_init, but
pyghmi has been changed for a long time to no longer have that
requirement.
 2017-10-20 15:18:03 -04:00
Jarrod Johnson
d26f6259a7 Add explicit error about certificate mismatch 
Before we were aborting without info.  Now explain the missing
information by including the error explaining why it is missing.
 2017-10-17 13:46:40 -04:00
Jarrod Johnson
bfc2b65e99 Support 'energy' sensor category 
Energy is distinct from power, categorize it as such.
 2017-10-17 11:07:54 -04:00
Jarrod Johnson
b2fdaeffe7 Add 'Power' to relevant categories for power 
While IPMI has no 'Power', some OEM sensors use that
type anyway.
 2017-10-13 16:08:59 -04:00
Jarrod Johnson
54f25cfa9e Implement access to other banks 
Provide backend support for other banks than primary.
 2017-08-16 17:30:18 -04:00
Jarrod Johnson
28ba40ccfa Replace SOL locknig with NullLock 
Again, eventlet monkeypatching of pyghmi seems to be messing up
use of RLocks...
 2017-07-31 10:05:50 -04:00
Jarrod Johnson
2ef549c2ae Correct the NullLock 
NullLock was not what util.protect was expecting, provide more function calls, all of which do nothing.
 2017-07-28 16:05:10 -04:00
Jarrod Johnson
9e44087047 Breakup logentries exceeding 65k 
The data length of a log entry must not exceed 65k.  If an attempt is
made to log that much, break it up and duplicate the records.  It may make
sense to indicate a continuation explicitly, but for now just extend.
 2017-07-28 15:53:07 -04:00
Jarrod Johnson
b5f016ad50 Provide support for removing update 
This allows for cancelling and cleanup after reporting.
 2017-07-26 14:10:04 -04:00
Jarrod Johnson
eb3f5a8bbc Provide update status through API 
Give visibility into the process.
 2017-07-26 11:39:27 -04:00
Jarrod Johnson
a40f015076 Phase 1 of firmware update support 
Provide ability to launch firmware updates.
Next will be ability to enumerate, monitor, and
delete.
 2017-07-25 17:01:32 -04:00
Jarrod Johnson
e0cc67f57a Implement Lenovo Thinksystem and discovery support 2017-06-20 14:56:24 -04:00
Jarrod Johnson
68037473d0 Replace login event mechanism 
Provide a different scheme that does not involve a wait(), if by chance
the flow dies without getting back to our thread.  wait() has no timeout
so this is a strategy to cope by making sure we hang for no longer than
3 minutes, which is well beyond any time a login should possibly take.
 2017-04-26 14:26:00 -04:00
Jarrod Johnson
6ab4d3c886 Add a recourse to drain queue 
While it may not have been possible in eventlet for this to happen,
strictly speaking if it were a thread, it could exit during check for
liveness and leave data on the queue.

To be careful, also drain the queue after all children dead.
 2017-04-25 16:48:35 -04:00
Jarrod Johnson
cf97bbe299 Rework ipmi worker management 
Provide a more concrete measurement of
children, rather than relying upon a
sentinel value on the queue.  It seems
that even using 'finally' didn't assure
that we always get that sentinel value
before a worker dies.  Sentinel value
still used to avoid a long wait in the
usual case.
 2017-04-25 15:27:29 -04:00
Jarrod Johnson
9e593f0554 Clean up nodepower 
nodepower boot and reset output was misleading.
 2017-02-15 10:46:27 -05:00
Jarrod Johnson
5ffc2c298b Correct mixup of parentheses and brackets 2017-01-04 09:56:31 -05:00
Jarrod Johnson
7bf8242aba Assign self.ipmicmd in case we are reusing an existing object that is already logged. 2017-01-04 09:45:19 -05:00