Discussion:
Corrupt database at first commit after a Fossil update
(too old to reply)
Zoltán Kócsi
2018-11-08 00:13:37 UTC
Permalink
I've updated Fossil from 1.29 to 2.7 on the server and on a client. The
database files on the server were left alone.

Then on a client, I did:

mkdir ~/ARM
cd ~/ARM
fossil clone https://<URL> <path_to_fossil_database>
fossil open <path_to_fossil_database>

All is fine. Or so it seems...

Then I modified a single file and attempted to commit. Note that both
the server and the client are now running 2.7, compiled from source.
Here's what happened:

***@linux-0ft9:~/ARM> fossil status
repository: /home/zoltan/fossil/armlib
local-root: /home/zoltan/ARM/
config-db: /home/zoltan/.fossil
checkout: 3fb6faec4bf580c7b97e4bdcc73774ce53f1f800 2018-11-07
01:34:08 UTC parent: f1c7e3ff8b960ffd30effc859468f1de269bbdbc
2018-11-06 23:29:22 UTC tags: trunk
comment: Changed the license to BSD (user: zoltan)
EDITED c/setjmp/setjmp.S

***@linux-0ft9:~/ARM> fossil commit -m "Fixed a bug in the M0
setjump"
Autosync: https://***@bendor.com.au/fossil/armlib.cgi
Round-trips: 1 Artifacts sent: 0 received: 0
Pull done, sent: 384 received: 383 ip: 203.16.199.2
New_Version: 62fba6495e1220468276520febd16cdfbf4056bc
Autosync: https://***@bendor.com.au/fossil/armlib.cgi
Round-trips: 1 Artifacts sent: 2 received: 0
Error: Database error: table plink has no column named baseid: {INSERT
OR IGNORE INTO plink(pid, cid, isprim, mtime, baseid)VALUES(2513, 2518,
1, 2458430.492028334, NULL)} Round-trips: 1 Artifacts sent: 2
received: 0 Sync done, sent: 19777 received: 409 ip: 203.16.199.2
Autosync failed.

According to the server's web interface, no change was done on the
server (last commit is still 3fb6faec4b). On the other hand, on the
client:

***@linux-0ft9:~/ARM> fossil status
repository: /home/zoltan/fossil/armlib
local-root: /home/zoltan/ARM/
config-db: /home/zoltan/.fossil
checkout: 62fba6495e1220468276520febd16cdfbf4056bc 2018-11-07
23:48:31 UTC
parent: 3fb6faec4bf580c7b97e4bdcc73774ce53f1f800 2018-11-07
01:34:08 UTC
tags: trunk
comment: Fixed a bug in the M0 setjump (user: zoltan)


Any clues as to what has happened and, more importantly, how to fix it,
would be very, very, very much appreciated. At the moment I don't dare
to touch either this or any of the other projects we have under Fossil
and of course work can't stop for too long :-(

I don't know whether that's important or not, but Fossil was configure-d
with the following flags:

--with-th1-docs --with-th1-hooks --with-tcl-stubs --json

Fuse and pledge are not installed on either the server or the client.

Thanks,

Zoltan
Stephan Beal
2018-11-08 00:20:32 UTC
Permalink
Post by Zoltán Kócsi
Error: Database error: table plink has no column named baseid: {INSERT
OR IGNORE INTO plink(pid, cid, isprim, mtime, baseid)VALUES(2513, 2518,
1, 2458430.492028334, NULL)} Round-trips: 1 Artifacts sent: 2
received: 0 Sync done, sent: 19777 received: 409 ip: 203.16.199.2
Autosync failed.
Version 1.29 was released in Aug. 2014, a few months before the baseid
field was added (Nov. 2014):

https://www.fossil-scm.org/index.html/info/5dd8b2dabead37a4

You probably just need to run 'rebuild' on the server once (or on a clone
of it, then re-upload that clone as the main copy).

You haven't lost anything - it's just a question of getting the schema back
in shape.
--
----- stephan beal
https://wanderinghorse.net/home/stephan/
<http://wanderinghorse.net/home/stephan/>
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
Zoltán Kócsi
2018-11-08 01:04:31 UTC
Permalink
On Thu, 8 Nov 2018 01:20:32 +0100
Post by Stephan Beal
You probably just need to run 'rebuild' on the server once (or on a
clone of it, then re-upload that clone as the main copy).
Yup, that was all.

Thanks a lot!!!

Best Regards,

Zoltan
Stephan Beal
2018-11-08 00:21:34 UTC
Permalink
Post by Zoltán Kócsi
I've updated Fossil from 1.29 to 2.7 on the server and on a client. The
database files on the server were left alone.
PS: this mailing list is no longer active. Please use the new forums for
future questions:

https://fossil-scm.org/forum
--
----- stephan beal
https://wanderinghorse.net/home/stephan/
<http://wanderinghorse.net/home/stephan/>
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
Zoltán Kócsi
2018-11-08 01:15:36 UTC
Permalink
On Thu, 8 Nov 2018 01:21:34 +0100
Post by Stephan Beal
PS: this mailing list is no longer active. Please use the new forums
https://fossil-scm.org/forum
Oh. That's really, really sad. I much prefer mailing lists to any web
based forum. Mind you, I'd choose NNTP over any web based thingy any
time. I guess it's because the colour (or lack thereof) of my hair.
Or that I have a mail/usenet client which is much more solid than any
browser and stores data on disk, so it also works off-line to a degree.
Well, times are a changin'...

Regards,

Zoltan
Stephan Beal
2018-11-08 01:24:40 UTC
Permalink
Post by Zoltán Kócsi
Oh. That's really, really sad. I much prefer mailing lists to any web
based forum.
Most of us do as well, but: bots and spammers are making it ever more
difficult to manage mailing lists. The forum can send you mail
notifications, but replying requires going through the forum. (i don't know
if there are plans to eventually support replies via email.) Unlike the
mailing list, the forum allows anonymous posts, but such posts are always
moderated before they're visible to the public. "So far" we haven't had any
spammers/bots on the forum, but that might be because it's not yet a big
enough target for them ;).

Well, times are a changin'...
We have the spammers and their bots to thank for that :/.
--
----- stephan beal
https://wanderinghorse.net/home/stephan/
<http://wanderinghorse.net/home/stephan/>
"Freedom is sloppy. But since tyranny's the only guaranteed byproduct of
those who insist on a perfect world, freedom will have to do." -- Bigby Wolf
Loading...