| *** altlogbot_3 has quit IRC | 02:12 | |
| *** altlogbot_2 has joined #starlingx | 02:14 | |
| *** dschoeffm has joined #starlingx | 04:26 | |
| *** dpenney_ has joined #starlingx | 04:51 | |
| *** dpenney has quit IRC | 04:53 | |
| *** cheng1 has quit IRC | 05:37 | |
| *** cheng1 has joined #starlingx | 05:37 | |
| *** anyrude10__ has joined #starlingx | 05:48 | |
| anyrude10__ | Hi Team, I was going through various Edge supporting projects. Wanted to have some idea about Airship vs StarlingX? AFAIK StarlingX also supports edge deployement in a production ready environment and Airship also has the same capabilities. Can someone please help me in this regard? | 05:48 |
|---|---|---|
| *** anyrude10__ has quit IRC | 07:30 | |
| *** jemangs has quit IRC | 08:07 | |
| *** jemangs has joined #starlingx | 08:10 | |
| *** ijolliffe has joined #starlingx | 13:00 | |
| *** dschoeffm has quit IRC | 13:14 | |
| sgw | Morning all | 15:08 |
| sgw | dirk: Morning, question about unresolvable issue with https://build.opensuse.org/package/show/Cloud:StarlingX:2.0/fm-rest-api, this worked earlier this week and is failing now due missing python2-amqp, I found python-amqp | 15:24 |
| sgw | marosales: I merged your first nfv, but then noticed the missing tis_patch_ver macro, I added a comment to the nfv review on how to fix that, please provide that fix and then we can try linkpac | 15:26 |
| marosales | morning team | 16:01 |
| marosales | I was trying to do that and it worked on my local package, but I couldn't push that change. I noticed you already have it here https://build.opensuse.org/project/prjconf/Cloud:StarlingX:2.0 isn't it applied to all the packages in that project? | 16:06 |
| marosales | I see there's this weird string lp150.1.1 for "tis_patch_version" in the packages generated https://build.opensuse.org/package/binaries/Cloud:StarlingX:2.0/nfv/openSUSE_Leap_15.0 | 16:11 |
| sgw | Yeah, I saw that just now, I am trying to do a local build | 16:11 |
| sgw | marosales: did you try doing a linkpac Cloud:StarlingX:2.0 nfv home:marcelarosalesj nfv-api-proxy | 16:14 |
| marosales | yes | 16:20 |
| sgw | marosales: interesting, when I build nfv locally I get 1.0-1, so the tis_patch_ver is getting propagated somehow. | 16:53 |
| marosales | yes, same for me, but only locally. Neither my OBS project has the tis_patch_ver substituted. dirk are you still here? do you know what's going on? | 17:32 |
| sgw | marosales: just reviewed your morning sumbission, did not see it earlier, sorry. | 18:05 |
| sgw | marosales: for the tis_patch_ver, did you do the "osc meta prjconf -e" and add the macro definition? | 18:06 |
| marosales | sgw: yes, I did that and worked locally | 18:11 |
| *** ijolliffe has quit IRC | 18:41 | |
| *** lcastell has joined #starlingx | 18:47 | |
| *** lcastell has quit IRC | 20:50 | |
| dirk | sgw: this was an ordering issue, in the prj path you had Stein last. the last project in the path is special, it is considered the base project which changes the build configuration. I swapped the order | 23:21 |
| sgw | dirk: thanks! | 23:23 |
| dirk | marosales: I don't see the problem atm. maybe it didn't rebuild? changing prjconf requires waiting a bit of time sometimes | 23:23 |
| dirk | sgw: build failure in https://build.opensuse.org/package/live_build_log/Cloud:StarlingX:2.0/fm-rest-api/openSUSE_Leap_15.0/x86_64 is a typical red hat thing, the path is called /etc/init.d in suse not /etc/rc.d. But you should really use systemd primitives instead | 23:24 |
| sgw | dirk: since your around for a few, I tried linkpac, but I think it broke when I did the osc sr Cloud:StarlingX | 23:24 |
| dirk | so "osc sr" is ~ similar to "osc copypac -e" (expand links, copy result) | 23:25 |
| dirk | so when you first linkpac and then sr, it will overwrite the destination with the expanded source | 23:25 |
| dirk | links are more of the type "just use whatever I point to, potentially apply the following modifications" | 23:26 |
| sgw | Dirk, I will look into the systemd primitives, so when we update the tarball, we need to update it in all locations if we use osc sr | 23:26 |
| dirk | so like a symlink+patch | 23:26 |
| dirk | while a submitrequest is "copy this revision, commit as new" | 23:27 |
| sgw | Since we want to do reviews I guess we go with copy instead of link. | 23:27 |
| sgw | I guess I should setup to build both SLE_15 and openSUSE_Leap to make sure we cover the differences in FHS | 23:28 |
| dirk | sgw: so when you have multiple OBS packages all from the same upstream tarball (eg. stx-fault-1.0.tar.gz here), a simple solution could be to create a package called stx-fault and just link that one under the other names | 23:29 |
| dirk | then you can just maintain all the spec files in one package and hence have only one source tarball | 23:29 |
| dirk | or use _multibuild | 23:30 |
| dirk | I think you experimented with that before, what was the reason for moving away from that? | 23:30 |
| dirk | SLE_15 and Leap15 are pretty similar, both have same path layout | 23:31 |
| sgw | I tested with _multibuild. but we wanted 1/package the linking your talking about is back to linkpac, but if we use osc sr that breaks the links, right? Or I have to pre-populate the Cloud:StarlingX with the linked packages. Also, we don't have an actual stx-fault package | 23:33 |
| sgw | an actual stx-fault specfile I mean | 23:34 |
| sgw | I created an empty one for the _multibuild | 23:34 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!