![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
I'm feeling a little stretched. Like butter that's been spread on too much toast.
Problem is, as a result, I'm also feeling like my ability to DO stuff is depleted.
There's kind of a satisfaction in knowing you're hosed, regardless. Because then you do what you can, and ... well if you blow it, well you did your best. You feel like there's no further need to push, beyond the limits.
I'm feeling that way now. I have until the end of the month to finish doing 'some stuff'. (SAN migration). It's doable, but ... very marginal. Our contingency time is almost gone, and we've still a lot outstanding that's ... well somewhat high risk.
We also have another project, that I will generously refer to as a cluster fuck from start to end. It's one of these Ultra screaming bitch high profile projects, that we have 'politics' piling the bullshit up. That's not a good start. However, it's also a house of cards infrastructure. 1 thing going 'ping' means it all falls apart. It's running on 5 year old servers, because the bastards were too tight to put it on 'new' stuff. Several of which only have one power supply - In servers, the bits that move fail. This is power supplies (well, their fans) and disks. Disks I can deal with, but if it's only one PSU ...
Oh, and the licenses for the software for the 10 servers it's on, costs about 100k each.
However, the real mammoth task for the next two weeks, is to build and 'make ready' as SAN attached Oracle veritas cluster.
Absolutely by the weekend of the 13th/14th. Because this political bullshit project is dependant on it. And they want 24/7, no downtime ever, because they're really important.
This project is also the one which I mentioned previously, that _has_ to have data coherency across all the servers, otherwise the backup is effectively corrupt.
The quantity of data involved takes _about_ 9 hours to copy. And they want it done in less than 30 minutes. Again, this is doable, as our new san can snapshot disks. Implementing this is going to be ... shall we say non-trivial. Especially as since project instantiation, I've not had a spare week or two to go on the 'how this stuff works' training course, that's really supposed to be a prerequisite.
Sometimes it's refreshing to know you're doomed to fail.
Problem is, as a result, I'm also feeling like my ability to DO stuff is depleted.
There's kind of a satisfaction in knowing you're hosed, regardless. Because then you do what you can, and ... well if you blow it, well you did your best. You feel like there's no further need to push, beyond the limits.
I'm feeling that way now. I have until the end of the month to finish doing 'some stuff'. (SAN migration). It's doable, but ... very marginal. Our contingency time is almost gone, and we've still a lot outstanding that's ... well somewhat high risk.
We also have another project, that I will generously refer to as a cluster fuck from start to end. It's one of these Ultra screaming bitch high profile projects, that we have 'politics' piling the bullshit up. That's not a good start. However, it's also a house of cards infrastructure. 1 thing going 'ping' means it all falls apart. It's running on 5 year old servers, because the bastards were too tight to put it on 'new' stuff. Several of which only have one power supply - In servers, the bits that move fail. This is power supplies (well, their fans) and disks. Disks I can deal with, but if it's only one PSU ...
Oh, and the licenses for the software for the 10 servers it's on, costs about 100k each.
However, the real mammoth task for the next two weeks, is to build and 'make ready' as SAN attached Oracle veritas cluster.
Absolutely by the weekend of the 13th/14th. Because this political bullshit project is dependant on it. And they want 24/7, no downtime ever, because they're really important.
This project is also the one which I mentioned previously, that _has_ to have data coherency across all the servers, otherwise the backup is effectively corrupt.
The quantity of data involved takes _about_ 9 hours to copy. And they want it done in less than 30 minutes. Again, this is doable, as our new san can snapshot disks. Implementing this is going to be ... shall we say non-trivial. Especially as since project instantiation, I've not had a spare week or two to go on the 'how this stuff works' training course, that's really supposed to be a prerequisite.
Sometimes it's refreshing to know you're doomed to fail.
