SequenceServer-PB: Non-functional (urgent)

[sdash-github]

User email on 20240730: Priority issue. "I'm having issues with the blast function of peanutbase. Earlier I tried to blast a sequence and the website was "BLAST-ing" for about 30 minutes. It never gave me results. Previously, when I blasted this particular sequence, it gave me results in less than a minute. My lab mate had a similar issue last week but she got an error message."

[sdash-github]

Used following seq for my test and it has already taken 20 mins and is still blast-ing while in LIS blast the result was in a few seconds.

65EGG atgcatgctttcttgtcaaatgaagggtgcaaaacatcatcactcatactagtcacttgg catgatagaacttcaaataactggctgtaatatcagaaatcctttctttgctgtttttct ggtcatggtttaagtcactctcatctgcaatgactatcattcattcattagattactgat tattcacttgctttgtagtagtgctaagtgctaactctttctttctacattggtaacagg agctttaacaacacaacaatgggagctggagggcgtgtcactaagattgaagctcaagag aagcctctttcaagggttccacattcaaaccctccattcagtgttggccaactcaagaaa gcaattccaccacattgctttgaacgttctcttttcatatcattctcctatgttgtctat gatctcttaatggcctacttactcttctacattgccaccacttatttccaccagcttcca tacccattttccttccttgcttggccaatctattgggccatccaaggctgcattctcact ggtgtttgggtgattgctcatgagtgtggccaccatgccttcagcaagtatcaacttgtt gatgacatggttggtttgatccttcactcttgtctattagtcccttatttctcatggaaa atcagccaccgccacctccactccaacaccggttccctcgaccgcgactaagtgtttgtc ctgaaaccaaaatcaaaggtatcacggtataacaagtacatgaacaatccactagagagg gctatttcccttttcatcacactcacactaggatggcccttgtacttggccttcaatgtt tctggcagaccctatgatagatttgcaagccactatgacccttatgctcccatatactct aacagggaaaggcttctaatttatgtctcaggactcagattcatctgtctttgctgtaac atatctgctatatcacatagcaactttgaaaggtttgggttgggtggtatgtgtttatgg ggtgccattgctcattgtgaatgggtttctagttaccataacctgtttgcaacacacaca tgcatcattgcctcactatgattcatccgaatgggactggttaagaggagcattggcaac agtggacagagattatgggatactgaataaggcatttcatcatataactgatacacatgt ggctcatcatttgttctcaacaatgccacattaccatgcaatggaagcaaccaatgcaat aaagccaatattgggtgattattaccaatttgatggcaccccattttacaaggcattgtg gagagaagccaaagagtgcctctatgtggagccagatgatggagcttctcagaagggtgt ttattggtacaagaacaagttctgatgcatagtgagagttggaaacgtgtgttaaattgg aaacttaagtaatttagtgacttgtaatagttatcaacaaaaat

[adf-ncgr]

looks like we've run out of space for holding results: sequenceserver_1 | 2024-07-18 05:39:56 - Errno::ENOSPC - No space left on device @ io_writev - /tmp/sequenceserver-138_hits.fa20240718-1-a8sppt (Errno::ENOSPC):

I've restarted the server and it seems to be working again but might also have to do some cleanup to prevent recurrence. Not sure offhand what the default for sequenceserver is in terms of how long results will be kept, let me know if you have thoughts on what would be reasonable.

[sdash-github]

I have asked the same users about a reasonable time to save the results. Let us wait for their response. "From your point of view what would be a reasonable period we should save a blast search result that you can still access without running the service again-- 3-4 days, a week, 2 weeks? We will really appreciate your input in this regard and will use that to periodically empty the space for newer results."

[adf-ncgr]

FYI per https://sequenceserver.com/doc/ it looks like:

BLAST results are stored for 30 days by default. You can change this by setting the :job_lifetime: key in configuration file (the value is specified in minutes).

Not completely clear to me whether it is working as advertised, however.

[sdash-github]

Both users reached back with these comments(20240731 emails): 1 week max. They were not aware that blast result is saved and they can retrieve it later. Almost always they revisit PB-BLAST with a new query and don't come back for the same blast result.

[adf-ncgr]

OK, I just updated the relevant config: https://github.com/legumeinfo/sequenceserver/commit/e4cfcbfe31af3d9b64c9e715591eeee51e803f40 but have not tested (and likely won't!)

[sdash-github]

I suppose you will later do the same in the ".....PeanutBase/sequenceserver/blob/master/lib/sequenceserver/config.rb" file? Change from existing value to num_jobs: 5, job_lifetime: 10_080, I suppose I can't update(get in) the peanutbase-sequenceserver, the only thing I can do is change the github repo via the web interface.

[adf-ncgr]

As far as I know we are not actually using that forked repository; I'm just using a branch on legumeinfo's fork to simplify my life (managing it as I do with the other genus-centric sequence servers). Not sure if there is a compelling reason to keep the peanutbase fork around.