Gentoo Archives: gentoo-dev

From: Jaco Kroon <jaco@××××××.za>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] rfc: switching default udev provider for new systems to udev
Date: Sat, 08 Aug 2020 21:38:45
Message-Id: 01679b97-2f98-8a1d-7800-e0d677a127ad@uls.co.za
In Reply to: Re: [gentoo-dev] rfc: switching default udev provider for new systems to udev by William Hubbs
1 -----BEGIN PGP SIGNED MESSAGE-----
2 Hash: SHA256
3
4 Hi,
5
6 On 2020/08/08 22:57, William Hubbs wrote:
7 > On Sat, Aug 08, 2020 at 09:17:20PM +0100, Roy Bamford wrote:
8 >> On 2020.08.08 19:51, William Hubbs wrote:
9 >>> All,
10 >>>
11 >>> I would like to propose that we switch the default udev provider on
12 >>> new
13 >>> systems from eudev to udev.
14 >>>
15 >>> This is not a lastrites, and it will not affect current systems since
16 >>> they have to migrate manually. Also, this change can be overridden at
17 >>> the profile level if some profile needs eudev (the last time I
18 >>> checked,
19 >>> this applies to non-glibc configurations).
20 >>>
21 >>> What do people think?
22 >>>
23 >>> Thanks,
24 >>>
25 >>> William
26 >>>
27 >>>
28 >>
29 >> William,
30 >>
31 >> With the declared aim from upstream of making udev inseparable from
32 >> systemd, its not something to be done lightly.
33 >> That's the entire reason that eudev was necessary.
34 >  
35 > Eudev never became necessary unless you are using a non-glibc system,
36 > and as I said, this can be handled in the profiles.
37 > Udev  runs completely fine without systemd, so I fail to see how eudev
38 > is necessary for most of Gentoo.
39
40 It actually works is enough reason for me.  Was forced to migrate a
41 bunch of systems not six months back from systemd-udev to eudev because
42 systemd-udev is absolutely terrible w.r.t. race conditions resulting in
43 lockups that kept forcing us into manual intervention situations. 
44 Mostly on systems with LVM.
45
46 I'm completely against the proposal.
47
48 >> I would want some convincing that it was not another step on the road
49 >> to Gentoo being assimilated by systemd.
50 >>
51 >> We had this discussion several years ago when the default became
52 >> eudev. What's changed?
53 >
54 > If systemd folks do make udev inseparable from systemd, then we would
55 > need eudev to be the default, but as I see it right now, there is not
56 > a case for it being the default.
57
58 Other than that it works and the systemd version does not.  Might be
59 configuration dependent, but I don't expect a default udev
60 configuration/system side to not cause LVM breakages when running
61 commands as simple as "lvs".  eudev in coparison just works.
62
63 >
64 > Another thing to consider is bus factor (eudev is maintained by one
65 > person primarily, so I have doubts about its viability as the default.
66
67 Yes, this is a problem.
68
69 Kind Regards,
70 Jaco
71
72 -----BEGIN PGP SIGNATURE-----
73
74 iQEzBAEBCAAdFiEEyyCUcKjG7P5BDam8CC3Esa/37p4FAl8vG1AACgkQCC3Esa/3
75 7p7Yvgf6Apoi1oCUKSyLEvH8fAEgbMIODULJAZx5+/C1dbROdjkWEzTTp3pNjWiQ
76 u8S2qz3xmh9QmKBwTAxB38U/gqXVRpF+xYfSF7K/CDUVcfAg5ViTL3W7YeJMPFNa
77 Jk8BgrarAc1Ln8OXCJ37Gf0eeuyBTsQQQ5qqubzNjdLBhrZegWY57gElrItE0Ywb
78 IjVBUO4QX3PSoOpZ5UlIo8Ioh+8ANXc/ADg7wASVQkd3dciyewZasZho/q6cNn6W
79 c44aMNFRTeiUfcK4+bpGMslq70y7D7JITkjkP+9e68e8wkh93L8fVs4BszBYEtUY
80 G6IXc4QtJ5Jf3bQRbyCnGcFYXrSLgg==
81 =rF5/
82 -----END PGP SIGNATURE-----

Replies