- Newest
- Most votes
- Most comments
Hi,
Thank you so much for asking your question.
AuroraDMLRejectedWriterFull is similar metrics as Aurora_fwd_replica_errors_session_limit which is stated in the document.
I will ask the document team to add AuroraDMLRejectedWriterFull metrics to the document to be more clear.
I hope this might help.
Hi,
It is the old name of renamed metrics AuroraDMLRejectedWriterFull
Then, what a full writer is and how to remedy is detailled here: https://repost.aws/questions/QUO2r2QFcCQG-F0egj97-xzg/rds-aurora-server-less-v2-writer-cpu-utilization-reaches-100-often
Best,
Didier
hello, We are aware that AuroraDMLRejectedWriterFull and AuroraDMLRejectedMasterFull are the same names depending on the version. What I'm curious about is what these indicators mean. I can't find an explanation of them on the page you linked.
Relevant content
- asked a year ago
Thank you for answer. Are you saying that this is the same error that occurs in the two cases of Aurora_fwd_replica_errors_session_limit in the link you shared? I think that when the writer is full, Aurora is currently using 128TB of capacity. What can we think of as the case of Too many forwarded statements in progress in general Aurora? If global or reader writes are not used, can I assume that the AuroraDMLRejectedMasterFull count increases only when the cluster's storage is full?
This is not related to storage size. Just mentioned on number of forwarded queries that get rejected due to exceeded the session on writer.
aha then In the case of the Aurora_fwd_replica_errors_session_limit indicator, as you mentioned, it may be affected by the number of forwarded queries. Under what circumstances can the AuroraDMLRejectedWriterFull metric occur if the add-on mentioned above is not used? Is this a situation where the connection has reached its maximum and additional creation cannot be made? Is storage size not taken into consideration with the relevant indicator?
Yes, storage size not taken into consideration. Both metrics can be used. I will raise this request to the team, meanwhile please use
Aurora_fwd_replica_errors_session_limitinstead to monitor forwarded query error.I don't use the forward function. There is a question about under what circumstances AuroraDMLRejectedWriterFull occurs. Can you share about the case before it was added to the doc?