You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Say you had a chunkserver or two missing, and a lot of data got replicated... now you re-add those chunkservers. Work has already been done and your data is now more secure... why undo it? Can you keep it without having to run a script to manually re-goal each chunk? And without having to wear your drives by increasing the goal on other files.
The text was updated successfully, but these errors were encountered:
The reason this is an issue, is that if a chunkserver goes down, chunks get replicated, which is good... but then when the chunkserver comes back up, the chunks get deleted- and while this is ongoing, throughput drops to below 1MB/s presumably because of RAM/CPU constraints on my chunkservers. Putting chunkservers that has gone down in maintenance mode when they are restored, temporarily alleviates the problem.
Say you had a chunkserver or two missing, and a lot of data got replicated... now you re-add those chunkservers. Work has already been done and your data is now more secure... why undo it? Can you keep it without having to run a script to manually re-goal each chunk? And without having to wear your drives by increasing the goal on other files.
The text was updated successfully, but these errors were encountered: