Invalidation trigger sets null status during save-progress action in launcher

[BDemps]

Summary: It seems that the catalog update is setting null in the Status field instead of "analysis in progress" when the user clicks the "save progress" button.

As of the past week or so, synspy-launcher (for sure in the mac version but maybe in the PC version, too...) has been having an issue where if a user hits "save progress" the actual result (as far as the launcher goes) is to put the file in the "complete" section (in my curator mode, it shows them as being "processed"). This is a new issue, so I'm not sure where it came from. It is hard to tell which exact files are affected since people have gotten assigned so many and I've been bouncing back the ones that I think are affected, but likely ones affected so far are (basically, any that have been submitted or saved between 12/21/2017 and 01/02/2018): SynImgZfZdu20171201A6A SynImgZfZdu20171114B3A SynImgZfDsy20171019B3A SynImgZfDsy20171208A3A and A6A SynImgZfZdu20171215B3A SynImgZfZdu20171208A3A

[karlcz]

@BDemps As I understand it, this is working for you on the Linux workstation? Can you do a test and confirm whether this issue affects you when you use the Mac OS launcher?

I want to isolate whether it is the Mac build or the user privilege level that makes a difference, i.e. student graders versus you with curator privileges on the server.

[karlcz]

Based on a few of these records I looked at, it seems like the "Status" field is getting set to null rather than "analysis in progress" in a request authenticated by the assigned student grader. So, I don't think it is the server-side automation causing the problematic records since that runs under a different robot credential.

We still need to isolate whether it happens with all launchers for all users or whether it is specific to a class of user or a particular build of the launcher.

[BDemps]

Could this have something to do with the fact that there is a new version of the synspy-launcher? None of the undergrads have this new version installed yet. If that would fix it, I'll send them the link as soon as possible

[BDemps]

Also, looks like the error also happens with the Linux version. If I do "save progress", it disappears. These are the two files I tested and it was the same for both! If you look in curator mode, you will see that they now say "processed", as though I hit the "complete", but I am certain that I hit "save progress".

SynImgZfDsy20170713B3A SynImgZfDsy20170713B6A

[karlcz]

OK, so this seems to confirm that it's a launcher bug affecting all builds and all users.

How critical is this? For the short term, do you think you can live with the workaround of completing a whole image stack in one session and saving it with the "complete" button?

[BDemps]

I think it's pretty critical. The undergrads are still on break for the week, and I think they are doing these in very small bites... I've had to bounce back a few times.

Why do you think that it could have happened all of a sudden? They've been using the same launcher for months now.

On Tue, Jan 2, 2018 at 3:07 PM, Karl Czajkowski notifications@github.com wrote:

OK, so this seems to confirm that it's a launcher bug affecting all builds and all users.

How critical is this? For the short term, do you think you can live with the workaround of completing a whole image stack in one session and saving it with the "complete" button?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_informatics-2Disi-2Dedu_synspy_issues_59-23issuecomment-2D354899635&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=vyxtQsOcnb8H62OHMJlLtCwxCrXTMlTajMGWts8yh9U&s=hlCqx-Wz7jcgJkiu4vt4obCjyjWkGcOuBu3YNVGNpug&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AbQSt0wPx-2D1Lq9eC-2DrElskfge7vvqx4cks5tGra3gaJpZM4RRFhd&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=vyxtQsOcnb8H62OHMJlLtCwxCrXTMlTajMGWts8yh9U&s=_bmYzyvT3XJRTZhd2QCtpEmFr-zar_8G21YpIJOeQA0&e= .

-- William Dempsey, Ph.D. Postdoctoral Scholar Molecular and Computational Biology Dana and David Dornsife College of Letters, Arts and Sciences University of Southern California wdempsey@usc.edu Lab #: 213-821-1818

[karlcz]

Didn't we do viewer2d changes for these users recently? To get those changes, they had to move to a newer build of the combined launcher-viewer2d for mac or windows.

But I don't think there have been server or client changes in the past couple weeks to easily explain a new bug. Is it possible it's been broken a lot longer but nobody noticed, e.g. because they usually completed a stack at once? Maybe that explains some of the previously incomplete stacks you found which prompted you to request the "bounce back" feature...?

[BDemps]

It's definitely new. There are several people in the lab who use their own laptops and they've definitely not done the full stacks at one time. There may be new changes to viewer2D but they haven't updated their synspy-launcher applications yet. It's still the older version.

On Tue, Jan 2, 2018 at 5:04 PM, Karl Czajkowski notifications@github.com wrote:

Didn't we do viewer2d changes for these users recently? To get those changes, they had to move to a newer build of the combined launcher-viewer2d for mac or windows.

But I don't think there have been server or client changes in the past couple weeks to easily explain a new bug. Is it possible it's been broken a lot longer but nobody noticed, e.g. because they usually completed a stack at once? Maybe that explains some of the previously incomplete stacks you found which prompted you to request the "bounce back" feature...?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_informatics-2Disi-2Dedu_synspy_issues_59-23issuecomment-2D354916754&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=Bd7HqU3H2VySpRtbYXgx1Lg2HXsZAWeJi25QTi2d8WI&s=4ANx-uwHfsazw2NN8ym0657Dh_rgpXCg9Vx5iZA8yMc&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AbQSt9tMm8iaPMMryXTQNGoFloSsvcTTks5tGtIcgaJpZM4RRFhd&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=Bd7HqU3H2VySpRtbYXgx1Lg2HXsZAWeJi25QTi2d8WI&s=KOA0a642342XxFuOz64nR3eA0QKJFd1sf3iE6eo_8P8&e= .

-- William Dempsey, Ph.D. Postdoctoral Scholar Molecular and Computational Biology Dana and David Dornsife College of Letters, Arts and Sciences University of Southern California wdempsey@usc.edu Lab #: 213-821-1818

[karlcz]

Actually, @BDemps I may have found a cause for this in a server-side DB function. I applied a small patch to try to correct that case. Can you re-test and see if this problem still affects you with your existing launcher?

[BDemps]

Yup, it's fixed now!

On Fri, Jan 5, 2018 at 11:57 AM, Karl Czajkowski notifications@github.com wrote:

Actually, @BDemps https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_bdemps&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=A7YWkhqIYZ5ScbDVzYRMvJxBeBn66T7SsuOi2UDD4Uw&s=ERKH-rRWPjJQdjDNj0bqHorrVoz54MzsRlIzvs-X5GQ&e= I may have found a cause for this in a server-side DB function. I applied a small patch to try to correct that case. Can you re-test and see if this problem still affects you with your existing launcher?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_informatics-2Disi-2Dedu_synspy_issues_59-23issuecomment-2D355650422&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=A7YWkhqIYZ5ScbDVzYRMvJxBeBn66T7SsuOi2UDD4Uw&s=UK7ahqKaSkD_QMF-d5SoHSbQbQIxAUT4RMMEQtkha7g&e=, or mute the thread https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AbQSt-5FsCxCwmm-5FED9BP0f1BBUlxT7Xaeks5tHn6egaJpZM4RRFhd&d=DwMFaQ&c=clK7kQUTWtAVEOVIgvi0NU5BOUHhpN0H8p7CSfnc_gI&r=Lssj6JKak_DaRnExGYcKnQ&m=A7YWkhqIYZ5ScbDVzYRMvJxBeBn66T7SsuOi2UDD4Uw&s=z3zvtySh3xim5PuSk4ejSu2OWYjJUQYZCDTadUuM9mM&e= .

-- William Dempsey, Ph.D. Postdoctoral Scholar Molecular and Computational Biology Dana and David Dornsife College of Letters, Arts and Sciences University of Southern California wdempsey@usc.edu Lab #: 213-821-1818

[karlcz]

Thanks for the confirmation.