Remote copy link down in hp 3par

[sumanth]

Hi Friends,
Please help me abot this alert in HP 3par 7400.

"Remote Copy Link Down Due To Error Or Missing Heartbeat"

how to resolve this issue and how we get this alers. please inform me.

[hdtvguy]

Welcome to my nightmare. Are you RCFC or RCIP? We get these message every day with RCIP, the 3par does a horrible job with RCIP and managing bandwidth. It expect perfect fixed bandwidth all the time and does not deal with any variations in bandwidth or latency very well. I also suspect it does a bad jb of handling RC when the pipe is full and congested. We have been working with HP since July of 2013, yes almost 18 months on these issues and are no closer than were were then.

There is a theory that we have worked with their engineering that the InformOS is too stupid to know what is going on beyond dumping data on the replication stack and thus if it does not see it's buffers deplete in an adequate time it assume a network issue and resends the data and thinks the link is down, even though it is not. The link shows up in IMC, the OS is just ignorant to what is going on and make bad assumptions which result in re transmits and wasted bandwidth when you can least afford the loss in throughput. RC for us is biggest Achilles heel of the array, we replicate hundred of volumes across 2 pairs of arrays and have around 300 RC Groups and live this pain every day.

[sumanth]

Hi Friend,
We are using RCFC. How to resolve this issue.

[Josh26]

We get these message every day with RCIP,

I'm glad I'm not the only one.

The question is, how do you get HP to stop calling you in the middle of the night when those alerts show up?

[Schmoog]

We get these message every day with RCIP,

I'm glad I'm not the only one.

The question is, how do you get HP to stop calling you in the middle of the night when those alerts show up?

I'll go ahead and throw a monkey wrench into the works and say I've been using rcip for over a year now and it's been spot on perfect. No errors except for very occasionally (5 or 6 times in the last 12 Months) taking longer than the sync period which in my case is set at a pretty aggressive 10 minutes.

Getting back to the op though, what does your inter site link look like? What devices are in the communications path? How much bandwidth?

[hdtvguy]

We get these message every day with RCIP,

I'm glad I'm not the only one.

The question is, how do you get HP to stop calling you in the middle of the night when those alerts show up?

I'll go ahead and throw a monkey wrench into the works and say I've been using rcip for over a year now and it's been spot on perfect. No errors except for very occasionally (5 or 6 times in the last 12 Months) taking longer than the sync period which in my case is set at a pretty aggressive 10 minutes.

Getting back to the op though, what does your inter site link look like? What devices are in the communications path? How much bandwidth?

I am guessing you have more bandwidth than you use most of the time then as when the WAN gets jammed you shoudl start to see these issues unless you have finely tuned the RC links with tunelinks commands and have the array throttle.

[hdtvguy]

We get these message every day with RCIP,

I'm glad I'm not the only one.

The question is, how do you get HP to stop calling you in the middle of the night when those alerts show up?

Not sure what level of support you have, but you can have your account team update the SSI info in the system with instructions to not call you outside certain hours for anything other than critical issues. It took us months of complaining and the support techs don;t always read the instructions, but it is better. What was worse was getting called in the middle of the night for a failed disk, really there is so much redundancy you are calling me at 2am for a failed disk.

3par support is still pretty bad, they have improved a bit since bring most of the US support back to the US, but their techs in support have very little experience.