Highlighted
1 Copper

SRDF reports inaccurate state for group

Hi All,

I have an SRDF/A group that gets into an odd state. It appears to just silently stop sending updates to R2 and R2 falls behind.

R2 is behind 14 hours. (From symrdf -g xxxx query -output xml_e)

<rdfa_time_r2_is_behind_r1>14:36:26</rdfa_time_r2_is_behind_r1>

<r2_data_is_consistent>True</r2_data_is_consistent>

   

So r2 is more than 14 hours behind. It is consistent. And strangely, there are no invalid tracks to send. This makes zero sense.

No invalid tracks (From symrdf -g xxxx query):

Total ------- ------- ------- -------

Track(s) 0 0 0       0

MB(s) 0.0 0.0 0.0     0.0


What we notice is that if we try to split the group, it will sometimes fail and we end up in suspended state but when we resume we find that now there are tons of invalid tracks and we may be 3TB+ behind. Sometimes the split succeeds but then we re-establish and right away we're 3TB+ behind, after only being split momentarily. This is the reason we believe the pre-split symrdf query is reporting incorrect information regarding invalid tracks.


Has anyone seen this issue? Any ideas to troubleshoot?

Thanks!

Vic


Labels (2)
0 Kudos
1 Reply
Highlighted
3 Zinc

Re: SRDF reports inaccurate state for group

This is SRDF/A MSC right?

You need to find ALL SE instances that are running the STORRDFD and check the RDFD logs on each of those hosts.

0 Kudos