I was told to report this on here:

I've got a list of ensembl IDs that's about 40k long so was trying to use biomaRt.

My first try was

mart <- useMart("ensembl")
mart <- useDataset("mmusculus_gene_ensembl", mart)
attributes <- c("ensembl_transcript_id", "transcript_appris", "ensembl_gene_id", "start_position", "end_position")
filters <- "ensembl_gene_id"
Output <- getBM(attributes = attributes, filters = filters, values=EnsemblIDList, mart=mart)

When I ran it it returns:

Error in getBM(attributes = attributes, filters = filters, values = EnsemblIDList,  : 
The query to the BioMart webservice returned an invalid result: biomaRt expected a character string of length 1. 
Please report this on the support site at http://support.bioconductor.org

However, after seeing https://support.bioconductor.org/p/105641/ I thought that maybe the issue was it was being directed to another server so I changed it to

mart<-useMart(biomart = "ENSEMBL_MART_ENSEMBL", 
              dataset = "mmusculus_gene_ensembl", 
              host = "www.ensembl.org",
              ensemblRedirect = FALSE)
attributes <- c("ensembl_transcript_id", "transcript_appris", "ensembl_gene_id", "start_position", "end_position")
filters <- "ensembl_gene_id"
Output <- getBM(attributes = attributes, filters = filters, values=EnsemblIDList, mart=mart)

Sadly, same issue. It gets a variable length through the process then dies; highest I've seen is 41% but generally less than 10%. Does anyone know what's going wrong?

The ensemblRedirect argument has been deprecated and doesn't do anything anymore. If you're using the latest version of biomaRt there should have been a message printed saying this.

To use one of the mirror sites you can do:

mart <- useEnsembl(biomart = "ensembl", 
                   dataset = "mmusculus_gene_ensembl", 
                   mirror = "useast")

Values for the mirror argument are: useast, uswest, asia.

This problem has been arising a lot recently, and I'm in contact with Ensembl to try and find a sustainable solution.