Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

BalancedShardsAllocator prematurely modifies `unassigned` shards list #3610

Closed
s1monw opened this Issue Sep 3, 2013 · 0 comments

Comments

Projects
None yet
1 participant
@s1monw
Copy link
Contributor

s1monw commented Sep 3, 2013

BalancedShardsAllocator modifies the unassigned shards list during initialization which is used by AllocationDeciders. Yet, premature modification can cause unintended throtteling of primary allocation which causes cluster to go in a red state until recovery / relocation has caught up to not throttle the allocation anymore. This is a very rare scenario that will only have a prominent effect on full cluster restarts or similar heavy weight operations.

@ghost ghost assigned s1monw Sep 3, 2013

@s1monw s1monw closed this in b39961b Sep 3, 2013

s1monw added a commit that referenced this issue Sep 3, 2013

Prevent ShardAllocator to modify the unassigned while running allocat…
…ions

The unassinged list is used to make allocation decisions but is currently
modified during allocation runs which causes primaries to be throttled
during allocation. If this happens newly allocated indices can be stalled
for a long time turning a cluster into a RED state if concurrent relocations
and / or recoveries are happening.

Closes #3610

mute pushed a commit to mute/elasticsearch that referenced this issue Jul 29, 2015

Prevent ShardAllocator to modify the unassigned while running allocat…
…ions

The unassinged list is used to make allocation decisions but is currently
modified during allocation runs which causes primaries to be throttled
during allocation. If this happens newly allocated indices can be stalled
for a long time turning a cluster into a RED state if concurrent relocations
and / or recoveries are happening.

Closes elastic#3610
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
You can’t perform that action at this time.