Bugzilla

(In reply to Jake Watkins [:dividehex] from comment #14)
> (In reply to Sarah Clements [:sclements] from comment #9)
> > :dividehex, would you be able to offer any insight or suggestions regarding the upgrading of prototype and stage instances and what would be involved?
> > 
> 
> I've gone down this road a few times and it plays out like this.  Request to increase space on rds, I increase rds storage size, management sees next months aws bill, request to downgrade back to original size, I downgrade size (nervously).  In 6 months to a year, rinse and repeat.  Sorry if reply comes across a little rough, I mean no harm.  I'm perfectly happy increasing the size of the rds instance storage BUT... the increased cost estimates must be approved up through management before proceeding.
> 
> Do you know what storage size you would like to target?  Are you also looking to upgrade the instance type on staging and prototype to match > prod?


Thanks for the backstory :) Unfortunately I do not know how much more we need for those instances, which makes this also not an ideal solution. I was more wanting to know whats involved since I haven't been historically. 

My preference at this point would be for the Treeherder and Perfherder team to come to a revised agreement about the max storage time for the performance_datum table (and for them to use active data as necessary, as Ionut has mentioned). I don't want to block on this bug, but the data retention for the perf team has been mentioned a few times in the past but never really reconciled. I'm hesitant to go through the whole song and dance of upgrading our RDS instances just for this purpose when I think the data retention issue should be addressed first.

(In reply to Jake Watkins [:dividehex] from comment #14)
> (In reply to Sarah Clements [:sclements] from comment #9)
> > :dividehex, would you be able to offer any insight or suggestions regarding the upgrading of prototype and stage instances and what would be involved?
> > 
> 
> I've gone down this road a few times and it plays out like this.  Request to increase space on rds, I increase rds storage size, management sees next months aws bill, request to downgrade back to original size, I downgrade size (nervously).  In 6 months to a year, rinse and repeat.  Sorry if reply comes across a little rough, I mean no harm.  I'm perfectly happy increasing the size of the rds instance storage BUT... the increased cost estimates must be approved up through management before proceeding.
> 
> Do you know what storage size you would like to target?  Are you also looking to upgrade the instance type on staging and prototype to match > prod?


Thanks for the backstory :) Unfortunately I do not know how much more we need for those instances, which makes this also not an ideal solution. I was more wanting to know whats involved since I haven't been historically. 

My preference at this point would be for the Treeherder and Perfherder team to come to a revised agreement about the max storage time for the performance_datum table (currently our second largest table at 19.2GB) and for them to use active data as necessary, as Ionut has mentioned above. I don't want to block on this bug, but the data retention for the perf team has been mentioned a few times in the past but never really reconciled. I'm hesitant to go through the whole song and dance of upgrading our RDS instances just for this purpose when I think the data retention issue should be addressed first (or at least a thorough investigation into whether active data is feasible).

(In reply to Jake Watkins [:dividehex] from comment #14)
> (In reply to Sarah Clements [:sclements] from comment #9)
> > :dividehex, would you be able to offer any insight or suggestions regarding the upgrading of prototype and stage instances and what would be involved?
> > 
> 
> I've gone down this road a few times and it plays out like this.  Request to increase space on rds, I increase rds storage size, management sees next months aws bill, request to downgrade back to original size, I downgrade size (nervously).  In 6 months to a year, rinse and repeat.  Sorry if reply comes across a little rough, I mean no harm.  I'm perfectly happy increasing the size of the rds instance storage BUT... the increased cost estimates must be approved up through management before proceeding.
> 
> Do you know what storage size you would like to target?  Are you also looking to upgrade the instance type on staging and prototype to match > prod?


Thanks for the backstory :) Unfortunately I do not know how much more we need for those instances, which makes this also not an ideal solution. I was more wanting to know whats involved since I haven't been historically. 

My preference at this point would be for the Treeherder and Perfherder team to come to a revised agreement about the max storage time for the performance_datum table (currently our second largest table at 19.2GB) and for them to use active data as necessary, as Ionut has mentioned above. I don't want to block on this bug, but the data retention for the perf team has been mentioned a few times in the past but never really reconciled. I'm hesitant to go through the whole song and dance of upgrading our RDS instances just for this purpose when I think the data retention issue should be addressed first (unless an argument can be made that the ROI to increase the RDS instances is worth the cost).

(In reply to Jake Watkins [:dividehex] from comment #14)
> (In reply to Sarah Clements [:sclements] from comment #9)
> > :dividehex, would you be able to offer any insight or suggestions regarding the upgrading of prototype and stage instances and what would be involved?
> > 
> 
> I've gone down this road a few times and it plays out like this.  Request to increase space on rds, I increase rds storage size, management sees next months aws bill, request to downgrade back to original size, I downgrade size (nervously).  In 6 months to a year, rinse and repeat.  Sorry if reply comes across a little rough, I mean no harm.  I'm perfectly happy increasing the size of the rds instance storage BUT... the increased cost estimates must be approved up through management before proceeding.
> 
> Do you know what storage size you would like to target?  Are you also looking to upgrade the instance type on staging and prototype to match > prod?


Thanks for the backstory :) Unfortunately I do not know how much more we need for those instances, which makes this also not an ideal solution. I was more wanting to know whats involved since I haven't been historically. 

My preference at this point would be for the Treeherder and Perfherder team to come to a revised agreement about the max storage time for the performance_datum table (currently our second largest table at 19.2GB) and for them to use active data as necessary, as Ionut has mentioned above. I don't want to block on this bug, but the data retention for the perf team has been mentioned a few times in the past but never really reconciled. I'm hesitant to go through the whole song and dance of upgrading our RDS instances just for this purpose when I think the data retention issue should be addressed first (unless perhaps an argument can be made that the ROI to increase the RDS instances is worth the cost?).