13 messages in com.selenic.mercurialhg clone via http broken?| From | Sent On | Attachments |
|---|---|---|
| Michal Kvasnica | Mar 9, 2006 5:02 am | |
| Thibault North | Mar 9, 2006 5:25 am | |
| Benoit Boissinot | Mar 9, 2006 9:15 am | |
| Benoit Boissinot | Mar 9, 2006 9:16 am | |
| Michal Kvasnica | Mar 10, 2006 6:13 am | |
| Thomas Arendsen Hein | Mar 10, 2006 8:49 am | |
| Michal Kvasnica | Mar 10, 2006 9:38 am | |
| Benoit Boissinot | Mar 10, 2006 9:59 am | |
| Michal Kvasnica | Mar 13, 2006 8:16 am | |
| Michal Kvasnica | Mar 24, 2006 7:47 am | |
| Thomas Arendsen Hein | Mar 24, 2006 8:21 am | |
| Thomas Arendsen Hein | Mar 24, 2006 9:56 am | |
| Thomas Arendsen Hein | Mar 24, 2006 12:55 pm |
| Subject: | hg clone via http broken? | |
|---|---|---|
| From: | Benoit Boissinot (bboi...@gmail.com) | |
| Date: | Mar 10, 2006 9:59:58 am | |
| List: | com.selenic.mercurial | |
On 3/10/06, Michal Kvasnica <kvas...@gmail.com> wrote:
Anyhow, I am still puzzled why a 4Mb cache worked with revision 35a05f177267 but does not work since 2e0a288ca93e.
Just to make sure it is not a demandload related problem, can you apply this patch on top of the broken 2e0a288ca93e and try to reproduce the problem? This reduces usage of demandload from 637 to 39 calls for cloning crew, and mpt is somewhat larger.
with rev 2e0a288ca93e:
$ hg clone http://control.ee.ethz.ch/~mpt/hg/mpt-25x/ requesting all changes adding changesets adding manifests adding file changes abort: premature EOF reading chunk (got 1717652 bytes, expected 1851712)! transaction abort! rollback completed
Can you capture the traceback with a network sniffer (i use tcpflow, but tcpdump can do it i think) ?
Thanks
Benoit