[Halld-offline] Hall D database server outage, Fri. March 22, 7:30 am

[Mark Ito]

hallddb is back up and serving databases. The outage was less than a 
minute. Twice the memory, twice the number of cores. We'll see if this 
makes a difference...

On 3/21/19 3:58 PM, Mark Ito wrote:
>
> Reminder: hallddb will have a service interruption tomorrow morning.
>
> On 3/20/19 4:34 PM, Mark Ito wrote:
>>
>> Folks,
>>
>> Our main offline database server, hallddb.jlab.org, will have a short 
>> outage on Friday morning, March 22, at 7:30 am. It is a virtual 
>> machine. The number of cores is being increased from 4 to 8. The 
>> memory is being expanded from 8 GB to 16 GB. This will hopefully help 
>> performance going forward, especially when many jobs start at the 
>> same time.
>>
>> The server should be down for a few minutes.
>>
>> Any farm job that *starts up* at this time and is using the MySQL 
>> interface to access CCDB or RCDB *will clearly encounter problems*. 
>> If the job has *gone past the point where it needs constants* from 
>> the database when 7:30 rolls around, it *should be ok* when the 
>> database server disappears. This has been tested, but not 
>> extensively. My judgment is that we live with a very few (if any) 
>> jobs having a problem rather than asking that all jobs be done before 
>> the upgrade occurs. Clearly if you are nervous about your jobs, do 
>> not run them during this time interval.
>>
>> We will likely be doing more to enhance database server performance. 
>> This particular upgrade is easy to do, causes minimal disruption, and 
>> should help so we are doing it first.
>>
>>   -- Mark
>>
>>
>
> _______________________________________________
> Halld-offline mailing list
> Halld-offline at jlab.org
> https://mailman.jlab.org/mailman/listinfo/halld-offline
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20190322/ec74c069/attachment-0002.html>