xCAT/confluent
2
0
Fork 0
mirror of https://github.com/xcat2/confluent.git synced 2025-01-16 04:39:16 +00:00
Code Issues Projects Releases Wiki Activity
1,685 Commits 39 Branches 64 Tags
Commit Graph

852 Commits

Author SHA1 Message Date
Jarrod Johnson
e9ba49a4aa Intercept another 'unexpected error' 
During a particularly hectic init, Invalid Session ID
may occur if a command is ran particularly early.  Intercept
and replace a more clean message.
 2018-10-12 15:46:54 -04:00
Jarrod Johnson
deb90fbca9 Fix trace on early console connect 
If the trace happens before tracelog is ready, just print the output
to the stdout log for now.
 2018-10-12 14:56:54 -04:00
Jarrod Johnson
3105b9b1f9 Significantly rework the collective startup behavior 
One, make the tracking bools enforce a lock to reduce confusion

Treat an initializing peer as failed, to avoid getting too fixated
on an uncertain target.

Make sure that no more than one follower is tried at a time by
killing before starting a new one, and syncing up the configmanager
state

Decline to act on an assimilation request if we are trying to connect
and also if the current leader asks us to connect and we already are.

Avoid calling get_leader while connecting, as that can cause a member
to decide to become a leader while trying to connect, by swapping
the reactions to the connect request.

Avoid trying to assimilate existing followers.

Fix some logging.
 2018-10-12 11:45:23 -04:00
Jarrod Johnson
f525c25ba6 Provide more verbose collective logging 
This helps understand the flow in practice of collective behavior.
 2018-10-11 15:15:11 -04:00
Jarrod Johnson
3012de1fe4 Prioritize deletion of transactioncount 
If the invalidation is incomplete, make sure that transactioncount
is invalidated first to avoid it being able to propogate through
a collective.
 2018-10-11 09:16:57 -04:00
Jarrod Johnson
be930fc076 Add missing subsystem marker from a collective log 2018-10-10 16:30:28 -04:00
Jarrod Johnson
2d0199a4e9 Wrap bdb deletion in same lock that sync itself uses 
If os.remove happens at a bad time, it causes an unfortunate behavior
in dbm.  Serialize this sort of operation to avoid the bad behavior.
 2018-10-10 15:24:55 -04:00
Jarrod Johnson
6b70a4322a Fix rollback 
The fix for the stale data introduced breaking clear rollback
Restore the behavior and make self._cfgstore a somewhat slower property
for now.
 2018-10-10 15:22:20 -04:00
Jarrod Johnson
6a784e3a1c Ensure sync is complete prior to leaving configmanager sync 
The initialization lock is meant to avoid collective and generic
initialization stepping on each other.  This is somewhat reduced in
efficacy if one has a sync running while the other is changing relevant
data.
 2018-10-10 14:49:33 -04:00
Jarrod Johnson
3b2b96a4cf Force fullsync if dead sync thread likely 
If the sync thread died previously, force the next sync to be full.
 2018-10-10 14:32:13 -04:00
Jarrod Johnson
32ddb33de3 Fix error when trying to do fullsync without globals yet 
If globals is missing, then do not break the sync trying to handle it
 2018-10-10 13:11:15 -04:00
Jarrod Johnson
b77ed8dbff Fix config sync on dead writer 
The sync thread can die without clearing syncrunning.  Make sure that
the thread is alive *and* that the thread has not indicated
intent to give up.
 2018-10-10 13:07:27 -04:00
Jarrod Johnson
d5c093a30d Provide fallback for unexpected reply in collective show 2018-10-10 09:46:01 -04:00
Jarrod Johnson
cf9d2a43e8 Revert "Provide fallback for unexpected reply in collective show" 
This reverts commit 2f566fb81ddfdd14b3b623ee6d1ff48d67e636b4.
 2018-10-10 09:44:06 -04:00
Jarrod Johnson
2f566fb81d Provide fallback for unexpected reply in collective show 2018-10-10 09:41:25 -04:00
Jarrod Johnson
1ee418392f Provide better error behavior on missing collective.manager 
collective.manager was a blanket response, make it per node and
only triggered by the bad nodes, not the rest.
 2018-10-09 15:44:17 -04:00
Jarrod Johnson
2a7eeb6e08 Fix missing argument in calling a function 2018-10-09 15:21:11 -04:00
Jarrod Johnson
5c83c78a90 Add warning on incompatible ssh key with SLES12 2018-10-09 14:44:06 -04:00
Jarrod Johnson
6a466b0100 Avoid proxy consoles generating proxy consoles 
When the client is a proxy term, disable ability to produce
proxy terminals.  This was wreaking havoc with client
count with ghosts and triggering output multiplication.
 2018-10-09 13:21:02 -04:00
Jarrod Johnson
5f46899358 Prevent clear_configuration from invaliding existing ConfigManager 
Clear out the existing dictionary instead of replacing it.

This prevents configmanager objects from being stuck.
 2018-10-08 16:51:58 -04:00
Jarrod Johnson
73c06fd25e Fix display of error on join of collective 2018-10-08 09:54:03 -04:00
Jarrod Johnson
8d9a082739 Provide better exceptions and propogate them to client on snmp 
When doing snmp, messages would always go to log only, even if the
user was at the confetty cli.  Give user access to knowing the error
impacting the query.
 2018-10-04 14:59:25 -04:00
Jarrod Johnson
32602fbba3 Provide interactive handling of key mismatch in ssh sessions 
Before, ssh would close without so much as a warning, fix this by
dealing with the key data.
 2018-10-04 10:23:55 -04:00
Jarrod Johnson
2f616d4586 Better error when collective.manager is set to something invalid 
If the collective.manager field does not correspond to any collective
manager, give a useful error rather than unexpected error.
 2018-10-03 16:23:20 -04:00
Jarrod Johnson
d86e1fc4eb Give the cfg init a lock 
Move collective manager and configmanager to share a configinitlock,
so that bad timings during internal initialization and collective
activity cannot interfere and produce corrupt database.

This became an issue with the fix for 'everything' disappearing.
 2018-10-02 10:17:44 -04:00
Jarrod Johnson
78a1741e0e Fix usage of check_quorum() 
It is not a boolean, it is exception driven.
 2018-10-01 16:02:16 -04:00
Jarrod Johnson
4329c1d388 Have collective start bail out if leader 
Leader should not relinquish if quorum, so don't bother in such
a case.
 2018-10-01 15:50:49 -04:00
Jarrod Johnson
b0b5493ff7 Cancel retry if we become leader 
If an instance is first to start, it's retry should be canceled
when other members prod it to become leader.
 2018-10-01 15:29:18 -04:00
Jarrod Johnson
326f56219b Fix /networking/macs/by-mac 
The module apimacmap was not correctly scoped.
 2018-10-01 14:40:02 -04:00
Jarrod Johnson
e098c0ba91 Fix missing tenant argument on user management function 
The tenant was omitted preventing those particular rpc calls from
working correctly.
 2018-10-01 14:04:03 -04:00
Jarrod Johnson
61e7c90ad1 Do not restart on intentional kill 
Additionally, add some output to help filter events log
 2018-10-01 10:32:55 -04:00
Jarrod Johnson
e57cdf9a7b Add more collective event log handling 
More detail to analyze how the collective membership is handled.
 2018-09-27 15:15:05 -04:00
Jarrod Johnson
10ce7a9de9 Add more logging to collective process 2018-09-27 10:51:06 -04:00
Jarrod Johnson
0724ad812b Add logging to the assimilation phase of collective 
When attempting assimilation, provide logging about the attempt.
 2018-09-27 10:51:01 -04:00
Jarrod Johnson
a3b0b0240d Abort assimilation attempt on non-member cleanly 
If a confluent instance has forgotten the collective, more cleanly
handle the situation, and abort the assimilation rather than assuming
the peer should be leader, unless txcount specifically is called out
as the reason.
 2018-09-27 10:50:54 -04:00
Jarrod Johnson
18bebde337 Disable gssapi in paramiko 
It is just plain broken, workaround by tanking calls to gssapi prior
to pulling in paramiko.
 2018-09-21 13:46:07 -04:00
Jarrod Johnson
f601032a66 Fix everything group missing if nodegroup created before node 
everything group was not making it to disk unless a node is created
first.  Correctly mark the need for disk sync to fix.
 2018-09-14 16:50:20 -04:00
Jarrod Johnson
db5f861dc5 Fix introduced typo in error message 2018-09-10 14:25:04 -04:00
Jarrod Johnson
d04be19ae5 Preferentially use a 'name' subfield as 'name' 
Pyghmi now may suggest a more useful name.  The component name
is unique, but 'name' can indicate the common name of things with
multiple instances.
 2018-09-07 14:37:02 -04:00
Jarrod Johnson
e7be24d478 Revert "Fix non-unique name for similar inventory items." 
This reverts commit 47a53a51e4bb625e28232cb8f792f2053b2e0b64.
 2018-09-07 11:44:01 -04:00
Jarrod Johnson
34b7abcb2d Change systemd unit to not have PIDFile 
systemctl restart *always* prints a worrying message
with pidfile.
 2018-09-07 11:27:43 -04:00
Jarrod Johnson
47a53a51e4 Fix non-unique name for similar inventory items. 2018-09-07 11:16:09 -04:00
Jarrod Johnson
b3bf6929df Add replacement logic for another generic variant 
In IMM, PCeGen3 x8 and similar is also possible.
 2018-09-06 16:16:26 -04:00
Jarrod Johnson
2a8d61ecf6 Enrich the less than useful 'Adapter' inventory items 
We can provide DNS provided info about such generic items to
make them look more fleshed out.
 2018-09-06 16:10:48 -04:00
Jarrod Johnson
cf3e9037ab Provide 'discovery.passwordrules' 
This provides an ability to designate the desired rules that
are applied in the wake of automatic discovery.  The most popular
would be 'expiry=no,loginfailures=0'
 2018-09-05 15:50:36 -04:00
Jarrod Johnson
03135543a6 Add 'switchuser' and 'switchpass' aliases 2018-09-05 13:51:19 -04:00
Jarrod Johnson
f92b1ed4a3 Implement ability to prompt for nodeattrib options. 
For certain attributes, notably passwords, it is sometimes desirable
to prompt interactively to help facilitate keeping such data out of
bash_history, screen sharing, and ps output.  -e enables this if the
user is aware of how to use 'read', -p is a quicker way to enable this.
 2018-09-04 09:38:01 -04:00
Jarrod Johnson
ba18b9936f Fix mistakes in previous commit 2018-08-29 15:15:34 -04:00
Jarrod Johnson
3b7ecd0095 Add ability to clear system configuration 
This provides a method to request the system firmware be restored to
factory defaults.
 2018-08-29 14:49:19 -04:00
Jarrod Johnson
8352007570 Limit to one active scan at a time 
Additionally, provide read access to rescan for discovery.
 2018-08-28 11:25:48 -04:00