16 Jun 2021 |
pdurbin | A nice presentation by Philipp @UiT/DataverseNO about CoreTrustSeal just now. Well done. | 12:48:42 |
17 Jun 2021 |
| hossamzalabany joined the room. | 08:37:19 |
hossamzalabany | Dear community, I have a quick question regarding Dataverse deployment, in the prerequisites manual, there is a recommended step to change the encryption method for localhost connection in pg_hba.conf to md5 instead sha256, is that obligatory? or it can work with both metods? | 09:08:15 |
hossamzalabany |
- for the postgresql setup*
| 09:08:52 |
poikilotherm | Can you link to the docs? | 11:10:01 |
poikilotherm | It sounds like a mistake - no one should be using MD5 these days | 11:10:17 |
poikilotherm | * hossamzalabany: Can you link to the docs? | 11:10:38 |
pdurbin | hossamzalabany: you're making me want to run a git blame to see when that docs was edited. Let me walk my kid to school and I'll take a look. | 11:37:35 |
pdurbin | hossamzalabany: is sha256 used in newer versions of postgres? I'm using 9.6 on my laptop and it still uses md5 (in pg_hba.conf). | 12:40:17 |
Don Sizemore | 9.6 and earlier versions of postgres always used md5. 13 supports scram-sha-256 assuming the JDBC driver is new enough to be compatible? | 13:15:39 |
18 Jun 2021 |
| Kacper Kowalik joined the room. | 14:22:50 |
Kacper Kowalik | pdurbin: hey! Has anything changed recently with dataverse.harvard.edu API access? It seems to be blocking a particular python client (urllib.request) | 14:23:38 |
Kacper Kowalik | from urllib.request import urlopen; urlopen("https://dataverse.harvard.edu/api/datasets/3035124") | 14:24:07 |
Kacper Kowalik | gives me 403 | 14:24:10 |
Kacper Kowalik | whereas requests.get("https://dataverse.harvard.edu/api/datasets/3035124") works just fine | 14:24:27 |
Kacper Kowalik | I assume it something along the lines of https://github.com/IQSS/dataverse/issues/5322 | 14:26:34 |
Kacper Kowalik | also searching by entityId:<id> doesn't seem to work anymore :( | 15:47:04 |
21 Jun 2021 |
pdurbin | Kacper Kowalik: | 13:56:17 |
pdurbin | Hey, sorry, we were out on Friday for Juneteenth. Sounds like you've got two issues and I'm afraid I don't immediately know what the cause of either could be. | 13:57:16 |
pdurbin | We have had some stability issues with Harvard Dataverse so it wouldn't surprise me if we now block some IP addresses here and there but that should change how the APIs work. | 13:58:10 |
pdurbin | I guess I would suggest testing a bit on https://demo.dataverse.org and emailing support@dataverse.harvard.edu about the issues you're seeing. | 13:59:00 |
Kacper Kowalik | pdurbin: for the first issue I have a workaround | 15:08:44 |
Kacper Kowalik | 2nd issue seems to be only affecting harvard's dv | 15:11:29 |
pdurbin | I'm glad you found a workaround but I sure hope APIs aren't changing. I don't know about that search issue either. We always index entityId. | 15:14:14 |
Kacper Kowalik | I sent a ticket to support for the 2nd one | 15:21:51 |
pdurbin | Ah. 305230. Thanks. | 15:26:55 |
pdurbin | Oh! https://dataverse.harvard.edu/file.xhtml?fileId=3040230 is my file! | 15:31:49 |
Kacper Kowalik | these are always your files. It may have something to do with the fact that I query your name when I look for testing data ;) | 15:32:43 |
pdurbin | 😀 Now we're talking about doing a Solr upgrade, which we should do at some point anyway. Hopefully it'll get fixed as part of that. | 15:58:21 |
pdurbin | Kacper Kowalik: thanks for brining this to our attention! | 15:58:54 |