Tomasz Chmielewski schrieb: > I just wanted to let you know that I just gad a tgtd segfault with a > 2008-06-09 git tree, after almost exactly 3 months of uptime. > > > tgtd[17500]: segfault at 8 rip 40ed9d rsp 7fff910a2750 error 6 > > > Is there some code in tgtd which tells it to segfault every 90 days? ;) > > This is a pretty busy server, with almost 200 initiators, so I wont try > to reproduce - I compiled today's revision, hopefully it'll work better. > > Thankfully, no initiator crashed (I think) after I restarted tgtd (there > was on process left). And this issue should be long fixed in git, anyway, I just had this one server not updated. BTW, what would be the best way to handle target crashes or updates, so that initiators are not interrupted? 1. Recommended HA solution: - set up two iSCSI targets; replicate everything between them with DRBD in dual-primary mode - on initiators, use multipathd - connect to both targets, set up multipath, and use /dev/mapper/S_multipath_device instead of a given iSCSI disk This way, you can stop one tgtd, and everything should still work, uninterrupted. 2. Poor man's solution: - you only have one iSCSI target server - how can you handle tgtd updates? - start two tgtd processes (i.e., on port 3260 and 3261), and connect your initiators to these two targets, and use multipath. Or? -- Tomasz Chmielewski http://wpkg.org -- To unsubscribe from this list: send the line "unsubscribe stgt" in the body of a message to majordomo at vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html |