-
Notifications
You must be signed in to change notification settings - Fork 1.1k
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
Data file size are different between nodes #4012
Comments
cc @amitanandaiyer , maybe similar to #3666 ? @zeinali0 Are your disks xfs? Also which version of yugabyte are you using? |
From the screenshot there's ~1200 tablets on each node. So ~1200 MANIFEST* files. So, like @bmatican suspects, if it is the issue #3666 (which has recently been fixed) -- that could explain about 1200*4MB = 4.8GB overhead on nodes where the yb-tserver was restarted (because of the unintended 4MB pre-allocation/fallocate call). This is a one-time cost per tablet in older release - and will be avoided in releases that include the fix for #3666. @zeinali0 -- can you look for some MANIFEST* files inside the data directory and compare the size reported by Thx! |
i'm using version 2.1.0.0. |
Yes it is duplicate. 66 byte file size and 4.0 MB on disk |
it increased to 10GB again |
Looks like the patch is in 2.1.2 yugabyte-db/src/yb/rocksdb/util/options.cc Line 264 in 50aaf8d
I tried locally and reports to 64KB. Can you retry @zeinali0 ? |
Thx for confirming @zeinali0 that it indeed is a duplicate of #3666. Like @ddorian mentioned, this issue is fixed in 2.1.2 release. Please see release notes here: https://github.com/yugabyte/yugabyte-db/releases/tag/v2.1.2 However, please note that the fix will not apply to tables already created. It will only apply to new tables created after the upgrade. |
The issue fixed after upgrading to 2.1.2.0. |
Why the tserver data size are different between nodes(logs size not counted)?
I have a 3 node universe with rf 3 but the size of data between nodes are too much different:
2 tserver are using about 10GB (number 1 & 3) while one of them using about 1.5GB(number 3)
while yb-admin UI is showing tablets are balanced between nodes
The text was updated successfully, but these errors were encountered: