Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/issues/221
yes, with this change
org.apache.bookkeeper.auth.TestAuth$LogCloseCallsBookieAuthProviderFactory.close
is called twice.
With or without this change BookieServer.shutdown will be
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
with this change they introduced multiplejournals/journaldirectories -
https://github.com/apache/bookkeeper/commit/123eccd435a4a96a9147ed4a24efbe9025fe79ba
---
If your project is set up for
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
and yes, I didn't add check for journaldirectories. Will add check for
journaldirectories also. Thanks for pointing that out.
---
If your project is set up for it, you can reply to
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
@sijie considering we have come to an understanding regarding the need and
purpose of this change, I'm keeping the logic as it is and just going to change
the config variable name
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
@sijie my bad, missed checking findbugs error after previous push. Fixed
them now and it passed.
Thanks.
---
If your project is set up for it, you can reply to this email and have
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
@sijie addressed your comments -
- move logic from BKShell to BookkeeperAdmin
- added functional testcases
- make DistributionSchedule and RoundRobinDistributionSchedule non-public
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/190
@sijie fixed tailing spaces
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
@sijie if you dont have anymore comments, this commit also should be good
to go.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/188
Thanks. I think this commit should be good to go.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
@sijie @eolivelli anymore comments/concerns?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/190
@sijie if you dont have anymore comments, I can fix the tailing spaces and
send new CR.
---
If your project is set up for it, you can reply to this email and have your
reply appear on
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
Yes @eolivelli.
I want the option to be available for users to choose whatever layout they
want and provide backward compatibility, but at the same time provide a config
to restrict
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/190
@sijie this is the fix we are having to deal with bookie start when
ledgerdirs are full.
---
If your project is set up for it, you can reply to this email and have your
reply appear on
GitHub user reddycharan opened a pull request:
https://github.com/apache/bookkeeper/pull/190
BOOKKEEPER-1034: Bookie start in RO when diskfull
When the disk is above threshold, Bookie goes to RO. If we have to restart
the
bookie, on the way back, bookie tries to create new
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
@sijie do you mean to say that remove the config and make it mandatory? I
dont think we can do that since it breaks backward compatibility and for
development and testing purpose there
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/189
1) Because of this we would get accurate disk usage/availability metrics if
multiple ledgerdirs are configured in the same drive. This is important for
DiskWeightBasedPlacementPolicy.
2
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/81
@sijie It should be ok for us, I don't have any in principle objection.
But will it possible to wait for one more day before you rebase. Let me
send my fix (ability to c
GitHub user reddycharan opened a pull request:
https://github.com/apache/bookkeeper/pull/189
BOOKKEEPER-1033: Handle DirsPartitionDuplication
- Provide configuration for allowDirsPartitionDuplication
- while calculating total disk metrics account Partition Duplication
You can
GitHub user reddycharan opened a pull request:
https://github.com/apache/bookkeeper/pull/188
BOOKKEEPER-1097: GC test when no WritableDirs
- Functional test validating that Compaction takes place even if there
are no writableledgerdir but there are ledgerdirs according to
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/81
@sijie ok, it is bit of an extreme case for it to happen (since
BookieStorageThreshold will be generally configured at safe levels) and even in
the case of eventuality (by misconfiguring or
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
@eolivelli I fixed it. It seems this time it has passed. Boxing/unboxing
checks is new thing to me. Thanks for the follow up.
---
If your project is set up for it, you can reply to this
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/81
your changes might not be required, because we are letting Bookie to start
in RO only mode when disks are full and within gcWaitTime period it will do gc
anyhow. Also, with your change it
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/81
How we handle when ledgerdirs are full when Bookie is started is little
different
- if isReadOnlyModeEnabled is true, then we try to bring up bookie in
ReadOnlyMode and provide ability to
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/186
@revans2 and @knusbaum FYI, LongHierarchicalLedgerManager needs this fix
for https://issues.apache.org/jira/browse/BOOKKEEPER-1096
---
If your project is set up for it, you can reply to
GitHub user reddycharan opened a pull request:
https://github.com/apache/bookkeeper/pull/186
BOOKKEEPER-1096: recursive znode delete
When ledger is deleted, along with leaf node
all the eligible branch nodes should be
deleted in ZooKeeper.
You can merge this pull request
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
@eolivelli @sijie
fixed rat check warnings you mentioned.
Yes 'decommissionbookie' bkshell command has to be executed in the node,
which is being decommissioned.
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/108
@sijie @eolivelli resolved merge conflicts and added more description to
the commit as we discussed in todays call. Thanks.
---
If your project is set up for it, you can reply to this
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/161
@jvrao @sijie we are using JMX not just for metrics (getters) but also for
invoking actions, like forcecompact, forceflush,.. It is the easiest way to do
out-of-band inter process
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/127
sure, will do. I was just waiting for full code review and comments, so
that I can address all of them at once.
Thanks,
Charan
---
If your project is set up for it, you can
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/81
@jvrao I haven't sent most of recent commits to community yet. one way or
the other they are dependent on https://github.com/apache/bookkeeper/pull/108.
I'm waiting for this com
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/108
@sijie its been quite long, can you please merge this commit. There are
multiple commits from our side which depend on this change and we are holding
them.
@jvrao @dlg99
---
If
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/108
@sijie @revans2
Addressed your comments and sent new commit, please take a look.
Thanks.
---
If your project is set up for it, you can reply to this email and have your
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/108
in the initially version we added code for Bookie to start in RO mode (if
RO mode is enabled) when all disks are full. But in the commit I created today
I removed that part, since it needs
Github user reddycharan commented on the issue:
https://github.com/apache/bookkeeper/pull/108
still working on this..will get back in couple of days
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does
GitHub user reddycharan opened a pull request:
https://github.com/apache/bookkeeper/pull/127
BOOKKEEPER-1028 and BOOKKEEPER-1029
BOOKKEEPER-1028: inc/excl opts listunderreplicated
- Introduce including and excluding BookieId options
for
35 matches
Mail list logo
