nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Imbalanced node storage consumption

JFMartin_ESI

In an SG GRID, there are 8 nodes, 4 are 5712, and 4 are 5760, all in the same storage pool, 2 buckets created, one for FabricPool and the other for Commvault backup destination. EC 2+1 used in the two buckets.

We noticed in recent days that one of the SG5712 node is reaching 93% while the other are around 57-70% used... SG5760 nodes are at 30% or a little bit less.

How is SG balancing the writes? Why is one node approaching 100% faster than the others?

Find more posts tagged with

Accepted answers

Amitra

HI @JFM ,

As per documentation, What are erasure coding schemes?

Erasure-coding scheme (k+m)	Minimum number of deployed sites	Recommended number of Storage Nodes at each site	Total recommended number of Storage Nodes	Site loss protection?	Storage overhead
2+1	3	3	9	Yes	50%

Its recommended to have 9 nodes to balance the Erasure Coding. Considering your current ILM and number of nodes, its expected one of the node may get over utilized.

Also, SG5760 are having higher disk capacity, hence, its expected their utilization will be lesser as compared to SG5712.

You may consider Rebalance using below documentation:

https://docs.netapp.com/us-en/storagegrid-117/expand/considerations-for-rebalancing-erasure-coded-data.html

https://docs.netapp.com/us-en/storagegrid-117/expand/rebalancing-erasure-coded-data-after-adding-storage-nodes.html

All comments

Amitra

HI @JFM ,

As per documentation, What are erasure coding schemes?

Erasure-coding scheme (k+m)	Minimum number of deployed sites	Recommended number of Storage Nodes at each site	Total recommended number of Storage Nodes	Site loss protection?	Storage overhead
2+1	3	3	9	Yes	50%

Its recommended to have 9 nodes to balance the Erasure Coding. Considering your current ILM and number of nodes, its expected one of the node may get over utilized.

Also, SG5760 are having higher disk capacity, hence, its expected their utilization will be lesser as compared to SG5712.

You may consider Rebalance using below documentation:

https://docs.netapp.com/us-en/storagegrid-117/expand/considerations-for-rebalancing-erasure-coded-data.html

https://docs.netapp.com/us-en/storagegrid-117/expand/rebalancing-erasure-coded-data-after-adding-storage-nodes.html

JFMartin_ESI

Thanks.

JFMartin_ESI

I forgot to mention that we are in a single site configuration, not a multisite configuration. Would the answer be the same?

Amitra

HI @JFM ,

Yes. The number of nodes will determine how balanced the ILM policy is along with their capacity. As recommended, you can try to perform rebalance to see if it helps the affected node.