[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[LDM #AKV-104910]: odd behaviour

Hi Karen,

I haven't heard of this before.

Are you collecting LDM metrics periodically via the command "ldmadmin 
addmetrics" and crontab(1)? It would be good to see the time-series plots.

Does your LDM report statistics back to us via rtstats(1)? If so, what's its 
hostname?

Does your LDM have a second REQUEST entry for the "NEXRAD2L2-BZIP2/KM" data?

> I've been having some odd issues that I was thinking were network related,
> but the more I look into them the more I wonder.
> 
> Basically, I'm wondering if anyone else has seen something like this, or if
> you have any idea what could be causing this to happen.
> 
> I've seen it happen on RedHat 5, RedHat 6 machines, and I've had it happen
> using ldm version from ldm6.6.5 (one of my long standing servers) all the
> way up to ldm 6.12.6.  I could try it on a newer ldm version, but I'm not
> sure it would make any difference.
> 
> The problem is with latencies in incoming data.  At times I will see a
> particular connection become very slow, and then it will clear itself.  Or
> sometimes restarting LDM will clear a latency problem.
> 
> For instance, I have a connection to IRADS for NEXRAD2 data.  I have split
> up my requests alphabetically to help to reduce latencies.  However, at
> times I will get a particular request that falls behind such as last night
> my request for "NEXRAD2L2-BZIP2/KM" seemd to slow down and then for a while
> I did not get any data for those letters. Specifically it slowed down
> 0700Z and didn't return to normal until 1000Z
> 
> During that time the 20+ other requests I have to the same IRADS server
> were fine, with no latencies.
> 
> For another example, I have a server here at NSSL that generates data and
> send it to a server at OU. We had puzzled over the high latencies with
> this particular data for quite a while. I had restarted the OU server
> multiple times, I split the requests up into multiple requests to try and
> resolve the problem, but nothing I did seemed to have any effect. Until I
> had some maintenance to do on the NSSL server. I made some modifications
> to it's pqact (which should not have affected the allows it had at all),
> and I restarted the LDM. Immediately the latencies went away and since
> then it's been fine.
> 
> I've seen a lot of similar instances of individual specific connections
> having latencies, and sometimes when going through logs I can see a lost
> connection followed by a reconnect that resolves the issues.
> 
> At any rate, if you have any ideas I would appreciate hearing them.

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: AKV-104910
Department: Support LDM
Priority: Normal
Status: Closed
===================
NOTE: All email exchanges with Unidata User Support are recorded in the Unidata 
inquiry tracking system and then made publicly available through the web.  If 
you do not want to have your interactions made available in this way, you must 
let us know in each email you send to us.