How does PKP LOCKSS work?

Software Support
1

We have 14 active and 3 archived journals and are running OJS 2.4.8.2. Currently we use Global LOCKSS and are considering whether or not to implement PKP LOCKSS. I understand LOCKSS on a theoretical basis but not the nuts and bolts (and I’m still not a tech expert), so I’m hoping someone can answer a few questions/confirm my understanding.

First, is it possible to run both Global LOCKSS and PKP LOCKSS at the same time? Could they both be run on the same journal? Could journal A use Global LOCKSS and journal B use PKP LOCKSS within the same installation?

Second, I have read that with PKP LOCKSS, content is automatically harvested when a new issue is published, so no human intervention is required, right? And the PLN harvests metadata plus galley content (PDFs, HTML, images–anything else? All other content files?) I’m sure it varies but is there some range for how long it takes harvested content to be processed?

Finally, a scenario (all too true, I’m afraid). Our OJS server was compromised and had to come down for a few days while a new one was built and content cleaned and moved over. We are trying to imagine what would have happened if we were using PKP LOCKSS.

From what I’ve read it sounds as though an outage would not automatically trigger access to journal content; but we would have to notify PKP LOCKSS of the outage and request … what? (It sounds as though we would make a request and someone would find the content and set it up on another server?) What else would we need to do? How quickly would we have public access to the content? And what would it look like to the public? (Obviously we are concerned with access as much as preservation, which may not be the same priority with the service.)

Thanks! Let me know if something isn’t clear.

Nancy

1 Like
2

Hi Nancy,

First, is it possible to run both Global LOCKSS and PKP LOCKSS at the same time? Could they both be run on the same journal? Could journal A use Global LOCKSS and journal B use PKP LOCKSS within the same installation?

Yes, it is possible, since the two plugins are unrelated (other than they both have “LOCKSS” in their names). Running both will not cause any problems.

Second, I have read that with PKP LOCKSS, content is automatically harvested when a new issue is published, so no human intervention is required, right?

Correct.

And the PLN harvests metadata plus galley content (PDFs, HTML, images–anything else? All other content files?)

The content that is preserved is generated by OJS’s Articles and Issues XML Plugin (more info on its output here). The intent is to be able to import this XML into an “access” OJS (see below for more info) after a trigger event. We also package up for preservation some basic metadata about the journal and its licensing terms, and some simple reports generated by the validation tools we run content through before committing it to the LOCKSS servers.

I’m sure it varies but is there some range for how long it takes harvested content to be processed?

It may take a few days (no more than three or four) for your content to be harvested, validated, and then deposited in the LOCKSS servers. It takes that long because the various steps required are scheduled to run daily.

From what I’ve read it sounds as though an outage would not automatically trigger access to journal content; but we would have to notify PKP LOCKSS of the outage and request … what? (It sounds as though we would make a request and someone would find the content and set it up on another server?) What else would we need to do?

If a journal manager contacts PKP indicating that her journal is being taken down permanently (or has gone down permanently with no hope of restoring it), we consider that a trigger event, and we would load the journal content into an “access node” for public consumption. Our plans are to have at least one access node (and perhaps only one, unless we can develop a good use case to have multiple access nodes), hosted by PKP.

How quickly would we have public access to the content? And what would it look like to the public? (Obviously we are concerned with access as much as preservation, which may not be the same priority with the service.)

We have not yet encountered a trigger event, so we don’t know exactly, but we would work with the journal staff to provide access as quickly as possible. I would say that once we are satisfied that the journal will no longer be accessible from its original source, we could have the content public in the access node within two weeks, possibly sooner.

As for what it would look like to the reading public, we are anticipating that all triggered journals will be hosted in a single OJS instance, the same way that many OJS users run multiple journals in a single instance.

Let me know if you have any other questions,

Mark
PKP Team

2 Likes
3

Where do I find the global lockss plugin? I found the plugin here: PKP PLN Plugin . But we want to be harvested globally, we got a global subscription anyway.
Thanks
Jan
gfgfg dfddf ffdfdfd

4

In OJS 2.4.x, the public LOCKSS interface is enabled in Journal Setup, step 2.6.

In OJS 3.x, this is found under Settings → Website → Archiving.

Note that checking the box does not automatically connect you with an existing public network. It only makes your content available, if the public network is aware of it. You will also need to contact the public network to make them aware of your journal.

5

Thanks. I knew about the enabling in step 2.6 but wasn’t aware to report it the network admin.

6

Hi @ctgraham
Does it mean that I must activate LOCKSS and then write https://www.lockss.org/ it to them so they can collect my magazine?

Screenshot_20180625_130719

Regards
Cristian

7

That is correct. After enabling the LOCKSS manifest within OJS, you must contact the network(s) you would like to harvest the journal, sending them the Publisher Manifest page.

8

Hello:

In ojs 3.1.1 works clockss? because I activate and in “Publisher Manifest” it does not list all the published numbers.

Regards