search

CDLUC3 / mrt-doc

Documentation and Information regarding the Merritt repository
8 stars 4 forks source link

Large object and dataset download email event recording recipients change #167

Closed elopatin-uc3 closed 4 years ago

elopatin-uc3 commented 4 years ago

Summary

Since the launch of Dryad, journal editors have expressed dismay about the additional recipients on the To line of the large object/dataset email messages being sent. This is due to the fact that the download feature is being used in conjunction with private peer review. The Dryad team has requested that we remove all internal CDL recipients from these messages, however we still need to be able to track that emails were sent successfully, and their content is correct.

Current recipient configuration for production is found here: https://github.com/CDLUC3/mrt-dashboard/blob/fb4f92c00e68e5a3af93846959a4aeda1df8f222/config/app_config.yml#L44

Ideas on how to address this issue:

Tasks

elopatin-uc3 commented 4 years ago

Sample log entry: Started GET "/asyncd/ark:%2F13030%2Fm52k1jd7/4?user_agent_email=ryaxley%40datadryad.org&userFriendly=true&lo sFrom=uc3%40ucop.edu&losSubject=Your+download+for+Urbanization+impacts+apex+predator+gene+flow+but+not+genet ic+diversity+across+an+urban-rural+divide+is+ready&losBody=The+dataset+that+you+have+requested+is+ready+for+ you+to+download.++Please+follow+the+link%0Abelow+to+download+your+file%3A%0A%0A++%3C%25%3D+%40dl_url+%25%3E% 0A%0APlease+note+that+this+link+will+expire+in+7+days+from+the+date+of+this+email.++The+content+is+stored%0A as+a+compressed+file+in+the+%22tar.gz%22+format.++For+many+operating+systems%2C+you+can+extract+the%0Afiles+ by+double-clicking+this+archive.++For+Windows+or+other+systems+you+might+require+a+utility%0Asuch+as+7-zip+% 28+see+http%3A%2F%2Fwww.7-zip.org%2F+%29+to+extract+the+files." for 172.30.22.60 at 2019-10-11 05:40:31 -070 0 Processing by LostorageController#direct as */* Parameters: {"user_agent_email"=>"ryaxley@datadryad.org", "userFriendly"=>"true", "losFrom"=>"uc3@ucop.edu ", "losSubject"=>"Your download for Urbanization impacts apex predator gene flow but not genetic diversity a cross an urban-rural divide is ready", "losBody"=>"The dataset that you have requested is ready for you to d ownload. Please follow the link\nbelow to download your file:\n\n <%= @dl_url %>\n\nPlease note that this link will expire in 7 days from the date of this email. The content is stored\nas a compressed file in the \"tar.gz\" format. For many operating systems, you can extract the\nfiles by double-clicking this archive. For Windows or other systems you might require a utility\nsuch as 7-zip ( see http://www.7-zip.org/ ) to ex tract the files.", "object"=>"ark:/13030/m52k1jd7", "version"=>"4"} Rendered inline template (0.4ms) Rendered text template (0.0ms) Completed 200 OK in 149ms (Views: 0.5ms | ActiveRecord: 1.6ms

marisastrong commented 4 years ago

would be great with this ticket to also update the documentation on redeploying the UI code https://confluence.ucop.edu/display/UC3/Merritt+UI+Production+Redeploy

elopatin-uc3 commented 4 years ago

@elopatin-uc3 to determine if immediate config change to remove CDL recipients is necessary as an initial way to address this.

elopatin-uc3 commented 4 years ago

Adding copy from an email that @dloy sent:

Hi Eric This is the error email sent when some type of error occurred during the async processing.

As you can see the To list is a bit outdated - the storage servers need the latest recipients - who would you like to include?

I thought this was handled as BCC - more research is needed.

Thanks David

-------- Forwarded Message -------- Subject: Exception: Your download for Data from: An ab initio electronic transport database for inorganic materials is ready Date: Tue, 5 Nov 2019 07:42:55 -0800 (PST) From: uc3@ucop.edu To: yasuyuki.masuda@murata.com, marisa.strong@ucop.edu, david.moles@ucop.edu, perry.willett@ucop.edu, jim.vanderveen@ucop.edu, eric.lopatin@ucop.edu

Unable to complete this requested action Customer service will be notified

The dataset that you have requested is ready for you to download. Please follow the link below to download your file:

http://merritt.cdlib.org/cloudcontainer/mrtstore1/901258329.tar.gz

Please note that this link will expire in 7 days from the date of this email. The content is stored as a compressed file in the "tar.gz" format. For many operating systems, you can extract the files by double-clicking this archive. For Windows or other systems you might require a utility such as 7-zip ( see http://www.7-zip.org/ ) to extract the files.

elopatin-uc3 commented 4 years ago

Initial pull request, just to remove inactive recipients from the To list: https://github.com/CDLUC3/mrt-dashboard/pull/37

Note that this work is now being tracked separately in https://github.com/CDLUC3/mrt-doc/issues/184

terrywbrady commented 4 years ago

If we want to use the BCC approach, here is a quick outline of the process I would follow to implement it.

Set bcc field through UI here.

https://github.com/CDLUC3/mrt-dashboard/blob/master/app/controllers/lostorage_controller.rb#L115

If a message object is found, parse the object in storage to determine if a bcc (or cc) field is present in message.

https://github.com/CDLUC3/mrt-store/blob/master/store-src/src/main/java/org/cdlib/mrt/store/action/AsyncContainerObject.java#L474

Set the BCC recipient using the following line as an example.

https://github.com/CDLUC3/mrt-store/blob/master/store-src/src/main/java/org/cdlib/mrt/store/action/AsyncContainerObject.java#L472

terrywbrady commented 4 years ago
elopatin-uc3 commented 4 years ago

Checked off first four tasks in the Summary. After discussion, use of only logging to record emails sent was not considered to be sufficient evidence of mail distribution. The option taken is to bcc a restricted set (two persons) of CDL recipients.