Hi,folks, Is there somebody has a experience to use MySQL replication to copy the mysql data from first box to second box? Can you share your experience of MySQL replication?
Thanks Zichao Wu On 6/15/06, John Lange <[EMAIL PROTECTED]> wrote: > On Thu, 2006-06-15 at 08:51 -0700, Shidan wrote: > > After reading up on new DNS features and some clarifications I've > > asked people, I have to agree SRV records are something I definitely > > have to implement, and John is totally right, but virtual ips are > > still a more full proof solution, there are times that SRV records > > won't work, using a cluster it will always work. I think that the very > > best solution is to do both, and neither are complex to implement ;) > > Definitely not trying to start an argument here; just curious; in what > situation would an SRV record not work but a cluster would? On the > contrary, I think there are several situations where SRV would work but > a cluster would not. Well just going by the voip-info entry you sent earlier: bad clients: - eyeBeam 1.1 3004w stamp 16863 uses first SRV, no failover, ignores ICMP (retransmissions) - eyeBeam 1.1 3010n stamp 19039 uses first SRV, no failover, ignores ICMP - ser 0.8.12 uses first SRV, no failover, ignores ICMP (retransmissions) - Asterisk CVS-HEAD-04/04/05-13:06:02 uses any SRV, no failover, ignores ICMP (retransmissions) - Windows Messenger 5.1 (5.1.0639), RTC API RTC/1.3.5369 uses first SRV, no failover, processes ICMP (no retransmissions) - Cisco 5300, IOS 12.3(10) uses first SRV, processes ICMP (1 retransmissions), failover to A record - Grandstream GXP-2000, 1.0.1.12 uses first SRV, no failover, ignores ICMP - Grandstream BR-102, 1.0.6.7 uses first SRV, no failover, ignores ICMP So it seems like anything that uses the traditional resolver libraries out there won't do failover. With time this will become less and less of an issue and probably things will happen sooner than later. But there seems to definitely be significant situations in the present > Clustering requires you to have the machines at the same physical > location (since they are sharing an IP). If something happens at your > location (fire, backhoe, internet provider routing issue etc.) both > nodes of your cluster will be down. Well true mostly, I would be scared if my VoIP carrier used only one ISP or relied on one router. > > SRV lets you have your servers geographically dispersed which is much > more reliable than a single location. This is definitely a good thing, the more redundancy the better, but more reliable than clustering, I don't see how. >And lets not forget that SRV also > lets you do load balancing at the same time as redundancy. Therefore, > unlike a fail-over cluster situation, many machines can be active and if > one goes down SRV transparently handles it. Same holds true for clustering > I more or less agree that a properly implement cluster would work just > as well in most situations you are likely to encounter but I will > respectfully disagree that clustering is as easy to setup and maintain > as a SRV DNS record. I'm still pretty convinced that clustering is a better solution, but DNS SRV is definitely something I will want to implement now. Cheers, Shidan --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
