SRA tests fail with missing accession

[nh13]

Description

See the travis build log: https://s3.amazonaws.com/archive.travis-ci.org/jobs/119821702/log.txt

   [testng] FAILED: ("SRR1063272", 0, 99, "SRR1063272.R.1", "ACTCGACATTCTGCCTTCGACCTATCTTTCTCCTCTCCCAGTCATCGCCCAGTAGAATTACCAGGCAATGAACCAGGGCCTTCCATCCCAACGGCACAGCA", "@@CDDBDFFBFHFIEEFGIGGHIEHIGIGGFGEGAFDHIIIIIGGGDFHII;=BF@FEHGIEEH?AHHFHFFFFDC5'5=?CC?ADCD@AC??9BDDCDB<", 86, "101M", "supercont2.1", 60, true, false)
   [testng] java.lang.IllegalArgumentException: Invalid SRA accession was passed to SRA reader: SRR1063272

This blocks #516 and likely any other PR that encounters these issues

Desired Actions

I would like to know why the test failed and what are the plans to ensure that the test does not fail in the future. This may not be a new issue, see #461, so I am looking for a detailed response and robust solution if this is a re-occurrence.

@a-nikitiuk could you take a look or route to the appropriate person?

Thank-you in advance to anyone who has the chance to take a look.

[a-nikitiuk]

NCBI had network connectivity issues today that resulted in bad database connections. Some of the issues have been cleared up already, but we continue to work on the problem. It has nothing to do with the NGS code and it is simply a network problem which resulted in a failure to resolve SRA accession that you saw in the tests

[nh13]

@a-nikitiuk that makes a lot more sense. Is there somewhere we could check the status of SRA network connectivity in case this happens again? A status page (ex https://status.github.com/) would be great to see the current and past uptime.

The error message should be improved. The message states that an invalid accession was given but really it was a problem querying SRA due to network issues. Perhaps something like "network connection timeout". @a-nikitiuk would you be able to update the error message? I can imagine a user who doesn't have their internet connected would be confused by such a message for example, in addition to when there is network connection problems at SRA. What do you think about modifying the error message?

[eitanbanks]

I agree that a better error message would be infinitely more helpful here.

[a-nikitiuk]

As for status page, you can look at https://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?view=announcement for any announcements in the future.

As for changing the error message, the current version of NGS API does not distinguish between network failure and a case when a provided accession does not exists. API only says, that it cannot resolve such accession to an SRA object, so we cannot change that message right now. However, we understand importance of a good error reporting and currently working on improvements in that area.

[tfenne]

@a-nikitiuk At least the error message should be changed to be not misleading, even if it is not currently possible to differentiate between the two. I.e. it could read could not resolve SRA accession; possible causes are an invalid SRA accession or that the SRA is currently unavailable. See https://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?view=announcement for SRA status.

[yfarjoun]

👍 for informative error messages.

On Mon, Apr 4, 2016 at 4:15 PM, Tim Fennell [email protected]
wrote:

@a-nikitiuk https://github.com/a-nikitiuk At least the error message
should be changed to be not misleading, even if it is not currently
possible to differentiate between the two. I.e. it could read could not
resolve SRA accession; possible causes are an invalid SRA accession or that
the SRA is currently unavailable. See
https://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?view=announcement for SRA
status.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub
#545 (comment)

[cmnbroad]

Can this be closed (fixed in #638 ?)

[a-nikitiuk]

Yes

[cmnbroad]

Closed via #638.