Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
/ 2004-09-13 11:14:30 +0800 \ P.V.Anthony: > Hi, > > I am having some problems with version 0.6.13 on 2.4.24. Whenever I try > to sync up, the secondary node will hang. > > Some details first. > > Linux: Fedora Core 1 > Kernel: 2.4.24 source from http://www.kernel.org > CPU: Intel P4 > Motherboard: Supermicro P4SCT+ > http://www.supermicro.com/products/motherboard/P4/E7210/P4SCT+.cfm > Drives: Sata Drives on Marvell Controllers > Network: Intel Gigabit > Raid: Software Raid 1 > FileSystem: ext3 > > > Here is want I did and please see if they are alright. > > 0. Waited for all software raid to sync up. > 1. Tried to "service drbd start" (version 0.6.13) on both servers > 2. cluster 2(secondary) will hang > 3. Tried drbd again but this time cluster 1(secondary) > 4. cluster 1 (secondary) will hang. So the seconadry will always hang. > 5. next I tried a old version of drbd 0.6.10 but same results > 6. next I recompiled the kernel without hi-mem IO (read in maillist) > 7. same results > 8. Then I tried 0.7.4 (solved the hi-mem IO problem) and it works well. > 9. Tried many restarts and it still works well. > 10. Next I tried openmosix with migration on and it works well. > > So I do not think it is a hardware problem. > > Please comment. Have I missed out any other thing? Is there any other > test I can do? > > Is there any patch for 0.6.13 so that it will not hang the kernel? I am > sorry for assuming that it is a 0.6.13 problem. I am a newbie and I > cannot think of any other reason. > > The reason why I want to use 0.6.13 is because I am using TKCluster from > http://www.solidrocktechnologies.com/?cmd=clustering > > Please do assist. why do you figure TKCluster is better suited for you than e.g. heartbeat? what exactly do you mean by "hang" ? still accessible? syslog messages? Lars Ellenberg -- please use the "List-Reply" function of your email client.